Electron beam focusing device using cylindrically symmetrical rotational crystal, and microfocus x-ray tube using the same

Abstract

To provide a charged particle focusing device using a cylindrically symmetrical rotational crystal capable of realizing the function of an electron diffraction lens.SOLUTION: An electron beam focusing device comprises a rotational crystal film having a cylindrically symmetrical rotational crystal with a thin, planar structure that has been cut out along a vertical plane direction with respect to a polar direction (Z) from a center (O) of a spherical crystal, and that has the thickness between 5 nm and 1 mm. When an incident electron beam passes through the rotational crystal film, the resulting strong scattering points converge at a focal position f located away from the film thickness of the rotational crystal film. Preferably, a polar angle (θ) of the rotational crystal film and a predetermined distance (r) approximately satisfy the following formula with respect to an incident direction (Zbeam) of the incident electron beam and an out-of-plane normal direction [hkl] of the rotational crystal film: r / f=tanθ (where f is a constant).SELECTED DRAWING: Figure 22C

Description

[Technical Field]

[0001] The present invention relates to an electron beam focusing device using cylindrically symmetric rotating crystals (CSRC), and a microfocus X-ray tube using the same. [Background technology]

[0002] Electron microscopy has always been an essential technology for observing the microscopic world and a crucial tool for testing and characterizing various materials. The development of the optical microscope enabled people to see cells, contributing to the development of accurate information in modern biology and medicine. The invention of the electron microscope enabled people to see atoms, enabling modern physics, materials science, and industry to make great strides over the past century. Advances in materials, such as synthetic fibers, building materials, superconducting materials, and aerospace materials, are relevant to daily life, energy issues, national defense, and the military. Advances in electron microscopy are the foundation for innovation in materials technology. Despite the technological maturity of microscopes, many problems remain to be solved, and technological innovation is still needed. Innovations in electron microscopy are expected to bring great value to a wide range of scientific and industrial fields.

[0003] The core of an electron microscope is the use of electromagnetic lenses to control an electron beam, focus it on a target material, and observe the emitted signal. Electromagnetic lenses differ in principle from optical lenses (diffraction gratings). While optical lenses (diffraction gratings) adjust the path of light by refracting or diffracting photons, electromagnetic lenses use electromagnetic fields to control negatively charged electrons. However, compared to optical lenses, electromagnetic lenses have several disadvantages.

[0004] First, electromagnetic lenses depend on a power source to operate, and using multiple lens sets in an electron microscope increases energy consumption. Second, the cost of high-precision, high-stability electromagnetic lenses is too high. Controlling electrons with an electromagnetic field is a relatively continuous process, and the stability of the electromagnetic field depends on the stability of the power supply that provides the field. Therefore, a high-precision, high-stability electromagnetic lens system requires a high-precision, high-stability power supply. Furthermore, because the strength of the electromagnetic field attenuates with distance and the frequency of the electromagnetic field also affects the control of the electron beam, the precision and stability of the electromagnetic lens vary depending on the wavelength of the electrons.

[0005] Third, electromagnetic lenses are large in size. Optical lenses require only a single piece of material and can be very small in volume, like a typical magnifying glass, and diffraction gratings can easily be a few centimeters or even smaller. Electromagnetic lenses require an electromagnetic field source and a sensor, whose volume often reaches tens of centimeters or even meters. Electron microscopes also often require the use of many sets of lenses, which often makes the electron microscope several meters in size, with some transmission electron microscopes reaching tens of meters in length.

[0006] Fourth, electromagnetic lenses inevitably produce chromatic aberration when focusing electron beams. Chromatic aberration occurs due to differences or fluctuations in the energy of the image-forming electrons. Because electrons move at different speeds within the lens's magnetic field, electrons scattered at a single point on the object surface cannot be focused. The same point on the image plane creates aberration. Therefore, electromagnetic lenses can only focus electron beams in the same direction, i.e., good electron beams. If the electron beams move in different directions or are not sufficiently monochromatic, the electromagnetic lens will not be able to focus the electron beams.

[0007] Fifth, it is difficult to eliminate aberrations in electromagnetic lenses. Optical lenses can simply correct aberrations by combining lenses to eliminate them. However, electromagnetic lenses operate on a different principle, so aberrations cannot be eliminated by combining lenses. This is also a theorem proposed by Scherzer, the father of aberration correction. From 1931 to 1995, it took scientists more than 60 years to finally design a spherical aberration corrector. This corrector was supplemented with a computer program with high-speed correction calculation capabilities to improve the effects of aberrations in electron microscope imaging.

[0008] If an electron diffraction lens similar to an optical lens (diffraction grating) that controls the focus of an electron beam by the refraction and diffraction of the material itself can be developed, the above problems could be fundamentally solved, and it is expected that electron microscopes that are more energy-efficient, more accurate, and lighter will be put into practical use, suitable for a new era. In this regard, as early as 1947, Boersch [1] proposed using the mean internal potential and material thickness to impart a phase shift to an electron beam. Hawkes et al. [2] also focused on achieving spherical aberration correction using charged foils, electrodes, and meshes. Bleloch et al. [3] attempted to fabricate thin phase-shifting films that functioned as electron lenses, but were unable to achieve practical applications. A more recent series of attempts by Shiloh et al. [4] used thin silicon nitride films machined with nanometer precision using focused ion beam milling. [Prior art documents] [Non-patent literature]

[0009] [Non-Patent Document 1] Boersch [J. Phys. Sci. 2 (1947) 615] [Non-patent document 2] Hawkes et.al. [Math. Phys. Eng. Sci. 367 (2009) 3637] [Non-patent document 3] Bleloch et. al. [Nature 394 (1998) 49] [Non-patent document 4] Shiloh et.al. Ultramicroscopy 189 (2018) 46 Summary of the Invention [Problem to be solved by the invention]

[0010] Diffraction requires a lattice spacing comparable to the wavelength. In the case of optical lattices, the wavelength of photons is typically several hundred nanometers, and the lattice spacing is on the order of microns, so it is easy to fabricate a diffraction grating of the corresponding size using etching techniques. In contrast, the wavelength of electrons is very short, on the order of a few angstroms, and the lattice constant of a typical crystal is also a few angstroms. Therefore, diffracting electrons requires atomic-level microfabrication capabilities, which is difficult to achieve in modern times. Therefore, to the inventor's knowledge, there are no publicly known documents related to electron diffraction lenses (lattices). [Means for solving the problem]

[0011] The objective of this invention is to design a special thin film material with a cylindrically symmetric rotational crystal structure and manipulate electrons passing through the thin film by utilizing the diffraction effect of the thin film material on electrons. The thin film material with a cylindrically symmetric rotational crystal structure is used to fabricate microdevices. These devices can control the trajectories of electrons, such as electron diffraction focusing lenses. Electron diffraction focusing lenses can replace conventional electromagnetic lenses in some specific application cases. Specifically, the inventors discovered that by adjusting experimental parameters during the process of depositing fine crystals in an amorphous film, it is possible to grow a rotated crystal with a very special crystal structure (different positions on the crystal film lead to different crystal rotation directions), leading to the invention of the present application. Here, the very special crystal structure refers to a crystal lattice with a crystal rotation direction connected within the fine crystal grains deposited in the amorphous film, and the atomic crystal planes of this rotated crystal material are arranged according to a special rule. The inventors discovered that this very special crystal structure can form a diffraction grating that can manipulate electron beams, leading to the invention of a charged particle focusing device using a cylindrically symmetric rotating crystal that realizes the function of an electron diffraction lens, and a microfocus X-ray tube using this crystal.

[0012] [1] The electron beam focusing device of the present invention is characterized in that it has a rotational crystal film having a cylindrically symmetric rotational crystal with a thin, planar structure cut from the center (O) of a spherical crystal along a plane perpendicular to the north pole direction (Z), as shown in Figure 22c, and the thickness of the thin, planar structure is 5 nm or more and 1 mm or less, and when an incident electron beam passes through the rotational crystal film, the resulting strong scattering point coincides at a focal position f that is farther away than the film thickness of the rotational crystal film.

[0013] [2] In the electron beam focusing device [1] of the present invention, preferably, an arbitrary point (R) of the thin planar structure is located at a predetermined distance (r) from the center (O) of the spherical crystal, at a position inclined by a polar angle (θ) with respect to the north pole direction (Z), and when expressed as a position inclined by an azimuthal angle (φ) with respect to the meridian connecting the north pole and south pole directions (Z) of the spherical crystal, The incident direction of the incident electron beam (Z beam ) and the polar angle (θ) and predetermined distance (r) of the rotational crystal film with respect to the out-of-plane normal direction [hkl] of the rotational crystal film preferably approximately satisfy the following formula: r / f=tanθ (f is a constant) [3] In the electron beam focusing device [2] of the present invention, it is preferable that the crystal orientation [hkl] of the cylindrically symmetric rotation crystal coincides with at least one of

[0111] ,

[0110] ,

[0100] ,

[0211] ,

[0121] , and

[0310] when the incident direction of the incident electron beam is

[0001] . [4] In the electron beam focusing device [1] of the present invention, preferably, the cylindrically symmetric rotation crystal does not have nanofiber regions with crystal boundaries, and the overall crystal orientation at adjacent positions in the azimuthal angle φ direction from the reference point O is continuously rotated. [5] In the electron beam focusing device [2] of the present invention, it is preferable that the locus of positions of the radius R where the polar angle θ changes continuously and the azimuthal angle φ does not change is linear, as shown in, for example, FIG. 15A. [6] In the electron beam focusing device [2] of the present invention, it is preferable that the locus of the position of the radius R when the azimuthal angle φ changes continuously and the polar angle θ does not change is approximately a circumference, as shown in, for example, Figure 15B. [7] In the electron beam focusing device [1] of the present invention, the thickness of the cylindrically symmetric rotation crystal is preferably 5 nm to 10 mm, more preferably 5 nm to 1 mm, and most preferably 5 nm to 60 nm, and the multilayer material of the rotation crystal film is preferably made of Si-doped InO and unavoidable impurities. If the thickness of the cylindrically symmetric rotation crystal exceeds 10 mm, the efficiency of focusing the electron beam decreases. This is because an electron beam lens using a cylindrically symmetric rotation crystal can focus an electron beam by utilizing the electron diffraction phenomenon, but the incident electron beam also scatters, reducing the focusing effect. If the thickness of the cylindrically symmetric rotation crystal is less than 5 nm, the thickness becomes too thin as a focusing lens, reducing its ability to focus the electron beam. [8] In the electron beam focusing device [1] of the present invention, it is preferable that the rotational crystal film has an amorphous film covering the cylindrically symmetric rotational crystal, and that the amorphous film contains a cylindrically symmetric rotational crystal other than the cylindrically symmetric rotational crystal.

[0014] [9] As shown in FIG. 22C, for example, the microfocus X-ray tube of the present invention is a microfocus X-ray tube comprising an electron gun and a metal target coated with the rotational crystal film described in [1] at a predetermined thickness, wherein the cylindrically symmetric rotational crystal coated on the metal target has a diameter of 1 μm or more and 20 μm or less, and the cylindrically symmetric rotational crystal coated on the metal target has a diameter of 1 μm or more and 20 μm or less, and an incident electron beam irradiated from the electron gun is incident on the cylindrically symmetric rotational crystal, and the cylindrically symmetric rotational crystal forms a bright spot of a Kikuchi diffraction pattern on the surface layer of the metal target, in a region where the bright spot is generated, and the incident electron beam is converted into X-rays, and the X-rays are projected onto a subject as a substantially point source of X-rays.

[10] In the microfocus X-ray tube [9] of the present invention, preferably, the surface layer of the metal target is a region of depth from the surface that includes a focal length f, such that the polar angle (θ) of the rotational crystal film and the predetermined distance (r) approximately satisfy the following equation: r / f=tanθ (f is the focal length)

[11] In the microfocus X-ray tube [9] of the present invention, preferably, the crystal orientation [hkl] of the cylindrically symmetric rotation crystal coincides with at least one of

[0111] ,

[0110] ,

[0100] ,

[0211] ,

[0121] , and

[0310] when the incident direction of the incident electron beam is

[0001] .

[12] In the microfocus X-ray tube of the present invention [9], preferably, the cylindrically symmetric rotation crystal is surrounded by an amorphous film, and The cylindrically symmetric rotation crystals are crystallized in different regions of the amorphous film, and the cylindrically symmetric rotation crystals crystallized in the different regions each exhibit different crystal orientations, The Kikuchi diffraction patterns are preferably present at positions (r, φ, θ) corresponding to the respective crystal orientations of the cylindrically symmetric rotation crystal.

[13] In the microfocus X-ray tube of the present invention [9], preferably, the focal distance from the rotating crystal film that focuses the incident electron beam (the distance between the interface between the thin film material and the target material and the position where the electron beam is most focused) is 50 nm to 1 mm, more preferably 500 nm to 100 μm, and most preferably 2 μm to 4 μm, and the focal size of the electron beam in the X-ray tube is 1 nm to 10 μm, more preferably 3 nm to 10 μm, and most preferably 12 nm to 500 nm. If the focal distance from the rotated crystal film that focuses the incident electron beam exceeds 1 mm, it becomes easier to control the range of change in the crystal orientation of the electron diffraction lens film, which is desirable from a manufacturing perspective, but this eliminates the need for a microfocus X-ray tube.If the focal distance from the rotated crystal film that focuses the incident electron beam is less than 50 nm, the greater the change in the orientation of the crystals in the rotated crystal film, the shorter the focal distance becomes, making it more difficult to control the range of change in the crystal orientation of the electron diffraction lens film, and making manufacturing difficult. If the focal spot size of the electron beam in the X-ray tube exceeds 10 μm, the quality of the X-rays will be poor and images with the required resolution will not be obtained.If the focal spot size of the electron beam in the X-ray tube is less than 1 nm, the focus of the electron beam spot will be too small, resulting in insufficient X-ray intensity and making it difficult to dissipate heat from the target material.

[14] In the fine focus X-ray tube of the present invention [9], the metal target is preferably at least one of Cr, Fe, Co, Ni, Cu, Mo, Ag, and W, or a material that may become an X-ray target material in the future.

[15] In the microfocus X-ray tube of the present invention [9], preferably, the electron diffraction lens operates as a microelectronic optical circuit to focus the incident electron beam. [Effects of the Invention]

[0015] According to the electron beam focusing device of the present invention, by appropriately selecting the crystal direction of the cylindrically symmetric crystal of revolution having a thin, planar structure, it is possible to manufacture an electronic operating device having different functions according to the crystal direction. Devices designed from cylindrically symmetric rotation crystals primarily manipulate the direction of electron emission through Kikuchi diffraction. The direction of Kikuchi diffraction is primarily controlled by the distribution of crystal planes in the sample and is less affected by the direction and energy of the incident electron beam. Therefore, devices designed from cylindrically symmetric rotation crystals do not require good directionality or monochromaticity of the incident electron beam and can be used with a wide range of electron sources. The device's functionality is controlled by how the crystal planes are distributed in the material. For example, electron diffraction lenses can focus electrons at any angle and energy distribution, and beam splitters can split an electron source into several electron beams for emission. The microfocus X-ray tube of the present invention is configured such that an incident electron beam is converted into X-rays in the region where bright spots of the Kikuchi diffraction pattern are generated on the surface layer of the metal target by the cylindrically symmetric rotation crystal, and the X-rays are projected onto the subject as a roughly X-ray point source. This allows for a structure that projects X-rays onto the subject with a focal length that is significantly shorter than that of conventional electromagnetic lenses. [Brief explanation of the drawings]

[0016] [Figure 1] FIG. 1 is an explanatory diagram of translational symmetry of a single crystal. [Figure 2] An atomic model of a silver-aluminum alloy quasicrystal with five-order rotational symmetry is shown. [Figure 3] An illustration of the arrangement of atoms in a spherical crystal, showing that point r(x,y,z) in space has atoms with spherical coordinates r(r,θ,φ). [Figure 4] This is an explanatory diagram of the cylindrically symmetric rotation crystal when it is considered as a thin slice cut from a spherical crystal along a direction perpendicular to the radius. [Figure 5] This shows the xy plane of a cylindrically symmetric rotation crystal, and the origin of the coordinate system is the center of the sphere of a virtual spherical crystal corresponding to the cylindrically symmetric rotation crystal. [Figure 6] The coordinate system shown has its origin at the center of the circle below the cylindrical base of a cylindrically symmetric rotation crystal. [Figure 7] The atomic distribution in the xz plane of a rotated crystal with cylindrical symmetry is shown. [Figure 8]The order of rotational symmetry at different radial positions in a cylindrically symmetric crystal of rotation is shown. (a) shows 5m rotational symmetry, and (b) shows 4m rotational symmetry. [Figure 9] 1A is a schematic diagram of the structure of a cylindrically symmetric rotation crystal and the direction of an electron wave vector according to an embodiment of the present invention, and FIG. 1B is a diagram illustrating the geometric relationship between spherical coordinates and rectangular coordinates in a cylindrically symmetric rotation crystal. [Figure 10A] (a) is a schematic diagram of raster scanning by a scanning electron microscope. [Figure 10B] (b) Schematic diagram of the incident electron beam and the rotated crystal island sample, (c) Schematic diagram of the change in the angle θ between the incident electron beam and the local crystal orientation of the electron-incident rotated crystal sample, and (d) Kikuchi pattern from a theoretical simulation of an SEM backscattered electron image in a two-dimensional rotated crystal. [Figure 11] (a) SEM image of a backscattered electron (BSE) detection of an InSiO film observed during crystallization of an amorphous film at 300°C. (b) EBSD image quality map (EBSD-Q) containing the distribution of crystal quality. (c) EBSD misorientation map (EBSD-M) and the distribution of misorientation, describing the relative orientation of two grains with respect to each other. (d)-(f) show inverse pole figures (IPFs) for the x-axis, z-axis, and z-axis, respectively, in degrees. [Figure 12] The crystal misorientation angle profiles measured along the arrows indicated at sites I to V within the rotated crystallization island. (a) shows the crystal misorientation angle profiles at sites I to V within the rotated crystallization island, and (b) shows the crystal misorientation angle profiles at each site I to V. [Figure 13] (a) shows a backscattered electron image, a secondary electron image, and a simulated image of a rotated crystal island whose central region is in the

[0222] crystal orientation; (b) shows a backscattered electron image, a secondary electron image, and a simulated image of a rotated crystal island whose central region is in the

[0400] crystal orientation; and (c) shows a backscattered electron image, a secondary electron image, and a simulated image of a rotated crystal island whose central region is in the

[0440] crystal orientation. [Figure 14]Detailed information on SEM images of a rotated crystal island with the central region in the

[0400] crystallographic direction, showing (a) the bandwidth, (b) the Bragg angle, (c) the interplanar spacing, (d) the lattice constant, (e) the strain, and (f) the distribution of the built-in geometrically necessary dislocation (GND) density. [Figure 15A] The distribution of crystal orientation (rotation angle) for an ideal cylindrically symmetric rotated crystal (a) and a real cylindrically symmetric rotated crystal (b) is shown, where the direction changes at each R position but the azimuth angle φ remains the same. [Figure 15B] The distribution of crystal orientations (rotation angles) for an ideal cylindrically symmetric rotated crystal (a) and a real cylindrically symmetric rotated crystal (b) is shown, where the [hkl] crystal direction at each R position changes but the polar angle θ remains the same. [Figure 16A] (a) is an overall perspective view showing one embodiment of a cylindrically symmetric rotated crystal of the present invention, (b) is a schematic diagram of the rotation of the crystal direction in an ideal rotated crystal island, (c) shows the radial crystal misorientation angle profile, and (d) shows the circumferential crystal misorientation angle profile. [Figure 16B] (e) is an overall perspective view of a conventional rotated crystal, (f) is a schematic diagram of the crystal direction rotation in a conventionally known rotated crystal island, (g) is the radial crystal misorientation angle profile, and (h) is the circumferential crystal misorientation angle profile. [Figure 17A] Schematic diagrams of Kikuchi bands formed in single-crystal thin-film materials by scattering of an electron beam. (A) is a schematic diagram when the incident angle of the electron beam coincides with the vertical axis of the single-crystal thin-film material, and (B) is a schematic diagram when the incident angle of the electron beam is tilted from the vertical axis of the single-crystal thin-film material. [Figure 17B] (c) is a schematic diagram of an ideal polycrystalline material capable of focusing an electron beam, and (d) is a schematic diagram of a rotating crystal capable of focusing an electron beam. [Figure 18] A flow chart showing the growth process of a rotated crystal, illustrating the deposition of an amorphous thin film of InSiO by magnetron sputtering. [Figure 19]FIG. 1 is a schematic diagram of a key portion of an electronic control device fabricated from a cylindrically symmetric crystal of revolution, with a single crystal island on the left and an array of single crystal islands on the right. [Figure 20] 1 is a flow chart showing a fabrication process of a rotating crystal membrane lens array. [Figure 21A] Flow diagram showing the growth process of a rotational crystal film lens, showing the deposition of an amorphous thin film of InSiO by magnetron sputtering. [Figure 21B] This is a flow chart showing the growth process of a rotational crystal film lens, showing the state in which a high-intensity electron beam is irradiated onto an amorphous film. [Figure 21C] This is a flow chart showing the growth process of a rotational crystal film lens, showing the transfer of a rotational crystallized InSiO film grown on a substrate to a TEM grid. [Figure 21D] Flowchart showing the growth process of a rotating crystal film lens, showing a schematic diagram of the SEM sample stage. [Figure 21E] A flow chart showing the growth process of a rotational crystal film lens shows the verification of transmission efficiency. [Figure 22A] FIG. 1 is a diagram illustrating the relationship between the size of the focal spot of the electron beam in a conventional X-ray tube and the degree of image blur. [Figure 22B] FIG. 1 is a diagram illustrating the relationship between the size of the electron beam focus and the degree of image blur in a conventional microfocus X-ray tube. [Figure 22C] 1 shows a focused electron beam coated microfocus X-ray tube illustrating one embodiment of the present invention. [Figure 23] Illustration of a rotated crystal island in the

[0111] direction: (a) is a cross-section of a rotated crystal island in the

[0111] direction; (b) is an enlarged cross-section of the crystalline region in the TEM image; and (c) is a fast Fourier transform (FFT) image of a selected region (marked by the white dashed rectangle in b) in the cross-section image. [Figure 24]SEM images of crystal islands of an InSiO cylindrically symmetric rotated crystal obtained with a backscattered electron detector. (a) is an image of multiple rotated crystal islands at an incident electron energy of 15 keV during the crystallization process. (b) is a backscattered electron image measured at tilt angles of -5°, 0°, and +5° of the

[0440] crystal island within the rectangular area in Figure 24(a). (c) is a backscattered electron image measured at incident electron energies of 5 keV, 15 keV, and 30 keV of the

[0440] crystal island within the rectangular area in Figure 24(a). [Figure 25A] The figures show the results of simulations of the Kikuchi diffraction patterns of an ideal InSiO cylindrically symmetric rotated crystal. (a) shows the simulated Kikuchi diffraction pattern of a crystal island of an ideal cylindrically symmetric rotated crystal. (b) shows the simulated Kikuchi patterns of the

[0440] crystal island in Figure 25(a) when the sample tilt angles are -5°, 0°, and +5°. (c) shows the simulated Kikuchi patterns of the

[0440] crystal island in Figure 25(a) when the incident electron energies are 5 keV, 15 keV, and 30 keV. [Figure 25B] The figure shows the simulated Kikuchi diffraction pattern of an ideal InSiO cylindrically symmetric rotation crystal. (d) shows the displacement of the Kikuchi pattern versus the tilt angle of the sample, and (e) shows the variation of the width of selected Kikuchi bands with the incident electron energy / wavelength. [Figure 26] A schematic diagram of an electron diffraction lens made from a cylindrically symmetric crystal of revolution with a single crystal island and a bulk electron gun in the center is shown. The figure shows the simulation results of the exit intensity cross section in the xy direction (propagation direction is z direction) after the electron beam emitted from the bulk electron gun enters the electron diffraction lens. [Figure 27] FIG. 1 shows the electron intensity distribution on the central axis of an electron diffraction lens made of a single-crystal island of an InSiO film with a thickness of 30 nm, when electrons are emitted from a bulk electron gun with an energy of 15 keV. [Figure 28] FIG. 1 illustrates an electron diffraction lens made of a cylindrically symmetric crystal of revolution with a polycrystalline island array. [Figure 29] FIG. 1 shows an electron beam splitter made of a cylindrically symmetric crystal of revolution. DETAILED DESCRIPTION OF THE INVENTION

[0017] 1: Concept of rotational crystals and cylindrically symmetric rotational crystals 1-1: Comparison of amorphous, quasicrystalline, single-crystal, and polycrystalline materials 1-2: Spherical crystals 1-3: Cylindrically symmetric rotation crystal 1-4: Cylindrically symmetric rotating crystal and electron beam 2: Kikuchi pattern principle and cylindrically symmetric rotation crystal (CSRC) sample 2-1: Principle of Kikuchi pattern in TEM images 2-2: Principle of Kikuchi pattern in SEM images 2-3: Kikuchi pattern using a cylindrically symmetric rotation crystal sample 2-4: SEM and EBSD images of a cylindrically symmetric rotational crystal sample 2-5: Analysis of SEM images 2-5-1: Symmetry analysis 2-5-2: Analysis of lattice constants 2-5-3: Analysis of reciprocal lattice vectors 2-5-4: Strain analysis 2-5-5: Analysis of built-in geometrically necessary dislocations (GND) 3: Cylindrically symmetric rotating crystal and its electron emission direction control 3-1: Two types of cylindrically symmetric rotation crystals 3-2: Difference between cylindrically symmetric rotating crystals and conventional rotating crystals 3-3: Control of electron emission direction from cylindrically symmetric rotating crystals 3-4: Electron beam focusing method for cylindrically symmetric crystal of revolution 4: Theoretical simulation method for electron diffraction from cylindrically symmetric crystals of revolution 5: Cylindrically symmetric rotating crystal production method 6: Device for electronic manipulation of thin films of cylindrically symmetric rotation crystals 7: A method for growing cylindrically symmetric rotating crystal focusing lenses in an array 8: High-throughput preparation and detection method using cylindrically symmetric rotating crystals 9: X-ray and fluorescence light source device using cylindrically symmetric rotation crystal film 10: Example 1 InSiO cylindrically symmetric rotation crystal 11: Example 2. Images generated by an electron beam interacting with a cylindrically symmetric crystal of revolution 12: Example 3 Cylindrically symmetric crystal of revolution for making electron diffraction lenses with single crystal islands 13: Example 4 Cylindrically symmetric rotating crystal for creating a multiple rotated crystal island array for electron diffraction lens 14: Example 5 Cylindrically symmetric crystal of rotation for making electron beam splitters

[0018] In order to make the objects, features and advantages of the present invention clearer and easier to understand, the following clearly and completely describes the embodiments of the present invention in conjunction with specific embodiments contained in the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, but are not all of them. Based on the following embodiments, all other embodiments obtained by those skilled in the art without requiring creative efforts, which do not change the basic principles of the claimed design, are included in the scope of protection of this application.

[0019] <1: Concept of rotation crystals and cylindrically symmetric rotation crystals> <1-1: Comparison of amorphous, quasicrystalline, single-crystal, and polycrystalline materials> A crystal is a structure in which atoms and molecules in a material are periodically arranged in three-dimensional space, and represents the most basic and essential characteristic of matter. Materials are classified into amorphous materials, quasicrystalline materials, and single-crystal materials according to the degree of atomic order in the material. Amorphous materials have the lowest degree of atomic order, while single-crystal materials have the highest degree of atomic order. A polycrystalline material is an aggregate of multiple crystal grains, each of which may be a quasicrystal or a single crystal, with the molecules and atoms inside arranged according to a certain rule. Multiple crystal grains of different sizes and shapes and with disordered orientation make up the polycrystalline structure.

[0020] Amorphous materials, also known as amorphous or glassy materials, belong to a class of rigid solids. Their constituent atoms and molecules are arranged in space without periodicity or translational symmetry, resulting in a breakdown of the long-range order of the crystalline state. Despite the presence of interconnections, the disordered nature of their internal atomic arrangements makes their structure difficult to describe in simple mathematical terms.

[0021] In single crystals, the fundamental symmetries of the crystal structure appear as translational and rotational symmetries, which can be described in mathematical terms. As shown in Figure 1, if an atom is located at a point r(x, y, z) in space, then three linearly uncorrelated basis vectors

number

number

number

[0022] Quasicrystals are ordered atomic arrangements that lack translational symmetry but have rotational symmetry. Three-dimensional quasicrystals with 5-fold, 8-fold, 10-fold, and 12-fold rotational symmetry have been identified. Figure 2 shows an atomic model of a silver-aluminum alloy quasicrystal with five-fold rotational symmetry. In addition to the materials mentioned above, there are also theoretical spherical crystals. Figure 3 shows the possible atomic structure of a spherical crystal. The black dots represent the smallest reproducible units, consisting of an atom or multiple atoms. In the following discussion, we will only consider the atomic case.

[0023] <1-2: Spherical crystals> As shown in Figure 3, a spherical crystal has an atom with spherical coordinates r (r, θ, φ) at a point r (x, y, z) in space. In other words, the arrangement of atoms in a spherical crystal has spherical symmetry. Atoms in a spherical crystal satisfy translational symmetry in any radial direction. In addition, when a rotation operation is performed, atoms in a spherical crystal at the same radial distance (r = r') rotate around the Y axis at an angle θ', which is expressed by the following equation, around the origin of the coordinate axes, i.e., the center of the sphere of the spherical crystal.

number

number

[0024] <1-3: Cylindrically symmetric rotation crystal> Using spherical crystals as a basis, we consider a new type of material: cylindrically symmetric rotation crystals. Cylindrically symmetric rotation crystals can be thought of as slices cut from a spherical crystal along a direction perpendicular to the radius, as shown in Figure 4. Depending on the atomic distribution, each cylindrically symmetric rotation crystal can be reduced to a virtual spherical crystal, i.e., each cylindrically symmetric rotation crystal is a part of a specific spherical crystal. Cylindrically symmetric rotation crystals have special translational and rotational symmetry in the arrangement of atoms and are cylindrical in shape. Cylindrically symmetric rotation crystals have spherical symmetry only in any radial direction of the corresponding virtual spherical crystal, with the center of symmetry of the corresponding virtual spherical crystal as the symmetry point.

[0025] As shown in Figure 5, the characteristics of a cylindrically symmetric rotation crystal in the xz plane are shown, with the center of the sphere of the virtual spherical crystal corresponding to the cylindrically symmetric rotation crystal as the coordinate origin. A cylindrically symmetric rotation crystal has an atom with spherical coordinates r(r,θ,φ) at a point r(x,y,z) in space. In other words, the arrangement of atoms in a cylindrically symmetric rotation crystal has spherical symmetry. Any atom in a cylindrically symmetric rotation crystal satisfies translational symmetry in the radial direction of the virtual spherical crystal after any reduction. In other words, the translational symmetry is equivalent to the following spherical coordinates.

number

[0026] Figure 6 shows a coordinate system with the origin at the center of the circle below the base of a cylindrically symmetric rotation crystal. If a cylindrically symmetric rotation crystal is cut from a corresponding spherical crystal along a direction perpendicular to the radius, it can be seen that a thin slice of a sufficiently thin thickness resembles a two-dimensional material. At this point, the symmetry of a cylindrically symmetric rotation crystal in the xy plane becomes more meaningful. Next, we will explain in detail the symmetry of a cylindrically symmetric rotation crystal in the xy plane. A cylindrically symmetric rotation crystal satisfies rotational symmetry with the axis of symmetry of the cylinder, i.e., the z-axis, as its rotation axis. The number of rotational symmetries varies depending on the radial position, and the number of rotational symmetries increases the further away from the axis.

[0027] Figure 7 shows the symmetry of a cylindrically symmetric rotation crystal in the xy plane, where φ1 represents the rotation angle of the first-level rotational symmetry around the origin O, φ2 represents the rotation angle of the second-level rotational symmetry, and φ3 represents the rotation angle of the third-level rotational symmetry. Let us now construct a coordinate system with the center point O of the XY plane of the rotating crystal as the origin, and each black dot represents an atom or the smallest unit (hereafter referred to as an atom). For the first-level rotational symmetry, six atoms are closest to the origin, and these six atoms are defined as the atoms of the first-level circle. The six atoms on the first-level circle satisfy 6-fold symmetry. That is, as shown by φ1 in Figure 7, each atom always coincides with another atom after rotating 60 degrees around the coordinate origin. Furthermore, since there are 12 atoms in the second-level circle, 12-fold symmetry is satisfied. That is, for the second-level rotational symmetry, each atom always coincides with another atom after rotating 30 degrees around the coordinate origin. Finally, there are 18 atoms in the third circle, satisfying 18-fold symmetry. For the third-level rotational symmetry, as shown by φ3 in Figure 7, each atom always coincides with another atom after rotating 20 degrees around the coordinate origin.

[0028] That is, a cylindrically symmetric rotation crystal has the z-axis as its rotation axis, and after rotating by a certain angle, the local atomic environment of the cylindrically symmetric rotation crystal does not change at the same distance from the center of the rotation axis, i.e., when r·sinθ is a constant value, and these angles can be expressed as φ=2π / m (m=1, 2, 3, ...). m is an arbitrary value that depends on the distance from the axial direction, i.e., the magnitude of r·sinθ. The greater the distance from the axial direction, i.e., the greater the r·sinθ, the greater the number of m. As shown in Figure 7, a cylindrically symmetric rotation crystal can draw many concentric circles depending on the arrangement of atoms, depending on the distance from the central atom O of the rotation axis. The rotation symmetry varies depending on the distance from the central atom of the rotation axis, for example, the mth circle has 6m times rotation symmetry. This property of a cylindrically symmetric rotation crystal is expressed by the fact that the arrangement of atoms in a cylindrically symmetric rotation crystal has an inherent rotation symmetry. In other words, in a cylindrically symmetric rotation crystal, if the atoms are at the same distance from the center of the rotation axis and the vector

number

[0029] Figure 8 shows the number of rotational symmetries at different radial positions of a cylindrically symmetric rotation crystal. The cylindrically symmetric rotation crystal in Figure 8(a) has an atom in the m-th ring and has five rotational symmetries. On the other hand, the cylindrically symmetric rotation crystal in Figure 8(b) has an atom in the m-th ring and therefore has 4m times more rotational symmetries for the atom in the m-th ring. Generally, materials are classified into amorphous materials, cylindrically symmetric rotation crystals (spherical crystals), quasicrystalline materials, and single crystal materials based on the degree of atomic order. Among them, cylindrically symmetric rotation crystals have a degree of atomic order intermediate between amorphous and quasicrystalline materials. Like quasicrystals and single crystals, cylindrically symmetric rotation crystals can form polycrystalline materials, which are cylindrically symmetric rotation crystals in which each crystal grain exhibits a different size, shape, and orientation.

[0030] <1-4: Cylindrically symmetric rotation crystal and electron beam> Fig. 9(a) is a schematic diagram of the structure of a single cylindrically symmetric rotation crystal and the direction of the electron wave vector, showing one embodiment of the present invention. Fig. 9(b) is a diagram explaining the geometric relationship between spherical coordinates and rectangular coordinates in a cylindrically symmetric rotation crystal. Here, taking into consideration the case where it is used as an electron diffraction lens, the electron beam e - The figure shows the case where the incident direction of the beam coincides with the z-axis. When a cylindrically symmetric rotation crystal is used for an electron diffraction lens, the electron beam e - It is advisable to select a coordinate system in which the incident direction of the electron beam and the z-axis coincide. For example, the spherical coordinates (r, θ, φ) and rectangular coordinates (x, y, z) for any position P in a single cylindrically symmetric rotation crystal have the relationship shown in Figures 9(a) and 9(b). That is, in a cylindrically symmetric rotation crystal 10, there is a reference point O with a center of symmetry in the crystal film. With this reference point O as the origin, rectangular coordinates x and y are assigned within the plane of the disc-shaped cylindrically symmetric rotation crystal for position P (x, y, z), and the z-axis is assigned in the out-of-plane direction. The electron beam e- is incident on a cylindrically symmetric crystal of revolution along the z-axis. On the other hand, the distance r in the spherical coordinates of the position P(r, θ, φ) represents the distance from the origin O. The polar angle θ in the spherical coordinates of the position P represents the distance from the electron beam e - The azimuth angle φ of the spherical coordinates of position P is defined in the range of 0≦θ≦2π, counterclockwise from the x-axis, by projecting position P onto the xy plane.

[0031] The base of the rectangular coordinate system (x, y, z) is expressed as follows:

number

number

number

[0032] In the cylindrically symmetric rotation crystal 10 of the present invention, a reference point O having a center of symmetry exists in the crystal film, and the crystal direction (vector) of the reference point O is designated as [hkl]. The center of symmetry of this crystal is not necessarily the geometric center of the cylindrically symmetric rotation crystal 10. The [hkl] crystal orientation at any position R of the crystal film is rotated with respect to the crystal orientation of the reference point O of the center of symmetry. The law of rotation is as follows. For convenience of explanation, we first establish a spherical coordinate system with the reference point O of the symmetry center as the origin and the [hkl] crystal direction at the reference point O as the z axis. This spherical coordinate system is used for the electron beam e when the cylindrically symmetric rotation crystal shown in Figure 9 is used for the electron diffraction lens. -This differs from the coordinate system in which the incident direction of the beam coincides with the z-axis. If the surface of the crystal film is defined as the xOy plane, the [hkl] crystal direction vector at a position R other than the reference point O can be expressed by the polar angle θ and the azimuthal angle φ in spherical coordinates (r, θ, φ) as shown in Figure 9(a). When the direction of OR remains the same but the scale changes, that is, when the reference point O is used as the center of a circle and the position R moves in the direction of radius r on the crystal film surface, only the polar angle θ changes continuously within the [hkl] crystal. 24(a) and 25(a), there may be multiple crystal islands discretely present in the crystal film as cylindrically symmetric rotation crystals 10. Each crystal island as cylindrically symmetric rotation crystals 10 has an [hkl] crystal orientation specific to each rotation crystal.

[0033] <2: Kikuchi pattern principle and cylindrically symmetric rotation crystal sample> <2-1: Principle of Kikuchi pattern in TEM images> Transmission Kikuchi Diffraction (TKD), also known as transmission-electron backscatter diffraction (t-EBSD), represents an advanced methodology for nanoscale orientation mapping. It facilitates the examination of the microstructural properties of thin specimens prepared for transmission electron microscopy (TEM) using a scanning electron microscope (SEM). This approach has attracted considerable attention in the analysis of nanocrystalline materials, oxides, superconductors, and a range of metal alloys. For a detailed discussion of Kikuchi patterns, see Williams, DB; Carter, CB, "Transmission Electron Microscopy." Springer, (2009) pp. 311-322, Chapter 19, "Kikuchi Diffraction." This document is incorporated herein by reference. In the TEM domain, Kikuchi lines are readily observable in diffraction patterns arising from sufficiently thick regions of a sample where multiple scattering occurs. In contrast to diffraction spots that flicker due to changes in crystal orientation, Kikuchi bands establish a structured orientation space with distinct intersections (called zones or poles) and paths connecting each intersection to the next. While discrete diffraction spots arise from the coherent scattering of the incident beam, the formation of Kikuchi bands is explained through a two-phase process involving incoherent scattering of the primary beam followed by coherent scattering of forward-scattered electrons. Furthermore, TKD has been used to analyze fine-grained ultramafic peridotite samples in the SEM. Sample preparation for TKD analysis can be performed using conventional methodologies applicable to TEM.

[0034] In TKD analysis, a thin foil sample is oriented perpendicular to the electron beam in a scanning electron microscope. The electron beam is precisely focused on a small area of ​​the sample, and the sample's crystal lattice diffracts the electrons that cross it. This diffraction pattern is then captured by a detector and subjected to analytical procedures to determine the sample's crystal orientation and microstructural characteristics [see Niessen, F., Burrows, A. & da Silva Fanta, AB A systematic comparison of on-axis and off-axis transmission Kikuchi diffraction. Ultramicroscopy186, 158-170 (2018)]. Convergent beam electron diffraction (CBED) is an additional TEM electron diffraction technique used to study Kikuchi bands. It involves the use of a converging or diverging electron beam (cone beam). First introduced by Kossel and Mollenstedt in 1939, CBED is useful for determining point and space group symmetries for crystal structure analysis [see Kossel, W. & Mollenstedt, G. Elektroneninterferenzen im konvergenten Bundel. Ann. Phys. 428, 113-140 (1939)].

[0035] <2-2: Principle of Kikuchi pattern in SEM images> Two modes of operation of a scanning electron microscope (SEM) are utilized to obtain comprehensive diffraction patterns. In the first mode, backscattered electrons are collected from various directions as they exit the sample surface, especially when the incident electrons strike the sample at a significant angle. In this mode, known as electron backscatter diffraction (EBSD), an incident electron beam enters the sample at a viewing angle and is scattered by atoms within it. The second mode, used to observe the full diffraction contrast, involves scanning the surface of a single crystal sample by changing the direction of incidence of the electron beam. This causes the angle around the crystal orientation to change continuously, causing the electron beam to diffract on the lattice at different positions on the sample surface, forming a Kikuchi pattern. This mode is called the electron channeling pattern (ECP).

[0036] <2-3: Kikuchi pattern using a cylindrically symmetric rotation crystal sample> While SEM raster scan mode is used to collect information about the sample in real space, ECP and EBSD modes are used to collect information about the sample in momentum space. In this study, we employed a unique sample configuration containing a cylindrically symmetric rotated crystal, whose orientation maintains cylindrical symmetry with respect to its central position while continuously rotating at local positions. This enabled us to observe the complete Kikuchi pattern in a single SEM image acquired in raster scan mode. The Kikuchi pattern observed in SEM images is created by a mechanism similar to that of electron channeling patterns (ECPs). Based on previously known approaches [Cheng, L., Ming, Y., & Ding, ZJ. Bohmian trajectory-bloch wave approach to dynamical simulation of electron diffraction in crystal. New J. Phys. 20, 113004 (2018); Cheng, L., & Ding, ZJ. Novel Quantum Trajectory Approaches to Simulation of Electron Backscatter Diffraction. eJ. Surf. Sci. Nanotech. 18, 121-125 (2020)], we use a dynamic approach to construct SEM patterns of rotated crystals by simulating the electron diffraction process in In2O3 single crystals at various incident directions. The simulation method will be described in a separate paper. Here, we theoretically reconstruct experimental SEM images using the observed crystal orientation distribution. In the simulation of this example, the fluctuation of the crystal rotation speed of each crystal island is ignored, and the average rotation speed, which is the uniform change of the angle between the incident direction of the electrons and the lattice of the In2O3 single crystal, is used.

[0037] Theoretical simulations show how rotated crystalline islands can generate perfect Kikuchi patterns in SEM images. Figure 10A (a) is a schematic diagram of raster scanning by a scanning electron microscope. Figure 10B (b) is a schematic diagram of the incident electron beam and the rotated crystal island sample. Figure 10B (c) is a schematic diagram of the change in the angle θ between the incident electron beam and the local crystal orientation of the electron-incident rotated crystal sample. Figure 10B (d) shows a theoretically simulated Kikuchi pattern for an SEM backscattered electron image of a two-dimensional rotated crystal. In Figure 10B(b), the thick downward arrow indicates the incident electron beam, the thin upward arrow indicates the rotation state of the crystal plane of the rotated crystal, and the thick jet-like cloud diffusing downward indicates the distribution of electron trajectories after the electron beam interacts with the material. Note that the thick and thin lines represent different electron probability densities. In Figure 10B(c), when the electron beam is scanned along a line passing through the center of the rotating crystal, θ is θ B From larger values ​​of θ B A and B in the figure change to a smaller value. B As shown in the bottom of Figure 10B(c), a change in θ is related to a change in the intensity of the backscattered signal at the corresponding position on the rotating crystal. Figure 10B(d) shows a theoretically simulated image of a local Kikuchi pattern in an SEM backscattered electron image of a two-dimensional rotated crystal.

[0038] Figure 10A(a) shows how an electron beam scans a cylindrically symmetric rotated crystal sample. Figures 10B(b)-(d) show how a diffraction pattern is generated by an electron beam on a cylindrically symmetric rotated crystal sample. Here, the local crystal orientation at different locations within the rotated crystal sample changes slowly and continuously, and the local crystal orientation at any location on the rotated crystal satisfies cylindrical symmetry with respect to the local crystal orientation at the center of the sample. As the electron beam scans the rotated crystal, the change in the angle θ between the incident electron beam and the local crystal orientation generates various backscattered signals collected by the SEM detector. When the sample is excited, backscattered electrons that satisfy the Bragg diffraction condition, 2d sin θ = λ, are diffracted by a specific set of crystal planes within the sample. This process forms two conical planes whose central axes are perpendicular to the crystal planes. The intersection of these two conical surfaces with the receiving screen creates an emission band known as the Kikuchi band, as shown in Figure 10B(d). At points A and B, the angle θ is θ B between A and B, θ<θ BTherefore, enhanced backscattering signals are detected at these positions, but before and after A and B, θ>θ B , the backscattered signal is reduced.

[0039] As shown in Figure 10B(d), the contrast in the 2D reconstruction of a typical 2D raster SEM survey across a rotated crystal includes contributions from several lattice planes together, and the resulting "electron channel map" displays contrast bands from all planes. In this map, the width of each band is equal to twice the Bragg angle (2θ) of the corresponding set of lattice planes, and the angle between different bands is equal to the angle between the sets of lattice planes from which those bands originate. In contrast, in conventional SEM mode, even if the electron beam is scanned perpendicular to the sample surface, incident electrons diffract from the atomic lattice with different directions at different impact locations. The resulting Kikuchi pattern is similar to that observed using EBSD.

[0040] <2-4: SEM and EBSD images of a cylindrically symmetric rotational crystal sample> Figure 11 illustrates SEM and EBSD images of a cylindrically symmetric crystal of rotation (CSRC) sample. Figure 11(a) shows an SEM image of the backscattered electron (BSE) detection of an InSiO film observed during crystallization of the amorphous film at 300 °C. Figure 11(b) shows an EBSD image quality map (EBSD-Q) containing the distribution of crystal quality. EBSD-Q is the sum of peaks detected by the Hough transform and represents the quality of the backscattered electron diffraction pattern. Figure 11(c) shows an EBSD misorientation map (EBSD-M) and the distribution of misorientation, which describes the relative orientation of two grains with respect to each other. Figure 11(d) shows an inverse pole figure (IPF) for the x-axis, in degrees. Figure 11(e) shows an inverse pole figure (IPF) for the y-axis, in degrees. Figure 11(f) shows an inverse pole figure (IPF) for the z-axis, in degrees.

[0041] To prepare the cylindrically symmetric rotational crystal sample, a DC magnetron sputtering system (Shibaura Mechatronics, CFS-4EP-LL i-miller) was used to deposit a 30 nm thick amorphous film on a sapphire substrate at room temperature. The sputtering target consisted mainly of In2O3 and SiO2, with a Si / In ratio of 2.3 at.% [1 wt.% of SiO2 / (In2O3 + SiO2)]. Dynamic crystallographic images along a precisely fixed observation area were obtained using in situ SEM observation by heating the InSiO2 film to 300 °C in the SEM chamber. The crystallization of InSiO2 during the annealing process is shown in Figure 11(a). As the temperature gradually increases, the amorphous film gradually crystallizes to form circular crystal islands with diameters of approximately 1–2 μm, exhibiting a distinct Kikuchi pattern. These rotated crystal islands grow from randomly distributed nucleation sites within the amorphous film and grow two-dimensionally outward. In most cases, the nucleation sites are isolated, resulting in the rotated crystal islands being separated from each other. However, in some cases, the presence of another crystallization point on an adjacent nucleus can cause the growth of the rotated crystal islands to stop, leading to the formation of rotated crystal islands that approach each other.This is based on the literature [Venables, JA & Spiller, GDT Nucleation and growth of thin films. Surface Mobilities on Solid Materials 341-404 (Springer, Boston, MA, 1983), Evans, JW, Thiel, PA & Bartelt, MC Morphological evolution during epitaxial thin film growth: Formation of 2D islands and 3D mounds. Surf. Sci. Rep. 61, 1-128 (2006), Shigeto, K., Kizu, T., Tsukagoshi, K., & Nabatame, T. Radial Interference Contrast in in-situ SEM Observation of Metal Oxide Semiconductor Film Crystallization. Microsc. Microanal. 23, 1512-1513 (2017), Gonzalez, D., Kelleher, JF, da Fonseca, JQ & Withers, PJ Macro This is consistent with the typical island nucleation and growth process in thin films reported in [and intergranular stress responses of austenitic stainless steel to 90 strain path changes. Mater. Sci. Eng. A, 546, 263-271 (2012).]. Figure 11(a) also shows that all these crystal islands exhibit distinct Kikuchi-like diffraction patterns in the SEM images. These crystal islands are observed to exhibit different Kikuchi patterns, suggesting different crystal orientations of these crystal islands. All the observed Kikuchi patterns belong to the cubic (bixbyite-type) In2O3 structure.

[0042] The observed Kikuchi patterns have several features, consistent with the preferred nucleation orientation during crystallization.

[0440] The preferred lattice orientation in oriented In2O3 is readily apparent from the observed Kikuchi patterns, the features of which are shown in Figure 11(a). This orientation can be derived from the Kikuchi pattern of the In2O3 crystal, and indices can be assigned to the Kikuchi bands using a spherical projection of the Kikuchi pattern. The nuclei always appear in pairs, exhibiting strong lattice curvature. This corresponds to the small contour distance between the Kikuchi bands (hkl) and (-hkl). The SEM image in Figure 11(a) shows two sets of (-22-2)-type planes, two sets of (-22-6)-type planes, one (00-4)-type plane, and one (4-40)-type plane, for a total of six major Kikuchi bands. Various bandwidths appear in the various sets of Kikuchi bands, forming bright "poles" as complex hexagons. Similarly, Kikuchi patterns with different characteristics of the

[0400] and

[0222] zone axes are also shown in Figure 11(a). These conditions are further confirmed by electron backscattering direction (EBSD) measurements, as shown in Figure 11(b) and (c). The normal EBSD plot (EBSD-ND) is inhomogeneous. Figure 11(c) shows the various orientations within the rotated crystal islands in an EBSD-ND map using black and white shading. Only one snowflake-like crystal "grain" is evident within each rotated crystal island, and no obvious crystal misalignment boundaries can be observed.

[0043] Figure 11(b) shows the EBSD quality (EBSD-Q). The EBSD-Q map shows that the crystalline islands have a snowflake-like mass distribution. The branching morphology of the crystalline clusters begins with a relatively uniform central region and decreases in mass as larger features develop. This suggests that it is the density difference between the crystalline and amorphous InSiO films that forms the rotated crystal orientation during the crystallization process. The EBSD misorientation (EBSD-M) map shown in Figure 11(c) shows that the misorientation distribution within each rotated crystal island has a snowflake outline pattern, and within that range, the misorientation is nearly identical, with a misorientation of about 1°. This can also be judged by the distribution of misorientations shown in the right panel of Figure 11(c). This result means that the crystal orientation indicated by any pixel in the image of these rotated crystal islands is nearly equal to the crystal orientation of the surrounding neighboring pixels. Inverse pole figures (IPFs) shown in Figures 11(d), 11(e), and 11(f) are plotted for the x-axis, y-axis, and z-axis, respectively, and are expressed in degrees.

[0044] Figure 12(a) shows five randomly selected directions (I–V) for three adjacent rotated crystal islands. Figure 12(b) provides quantitative information on the rotation of the local crystal orientation in Figure 12(a), plotting the change in the crystal orientation of the rotated islands in the five selected directions. Although the three adjacent rotated islands differ in size and crystal orientation, the rotation rates of the five randomly selected directions are nearly equal, despite the differences in the rotation speed and crystal orientation of the rotated islands. The average rotation rate of the crystal rotation angle is obtained by linear fit. In this rotated crystal sample, the average crystal orientation rotation rate of the three adjacent rotated crystal islands is 16.62° / μm. Note that the average crystal grain size of the three adjacent rotated crystal islands is approximately 1 μm.

[0045] <2-5: Analysis of SEM images> The Kikuchi pattern observed in secondary electron (SE) images is thought to result from electron diffraction on the periodic atomic lattice of the sample. It is interesting to simultaneously observe diffraction contrast and morphology contrast in SEM images. Here, an InSiO film is first annealed in air to 250°C, then transferred to the SEM chamber and further annealed to 300°C. Following this, small spherical protrusions with a radius of approximately 20–30 nm can be observed in SEM images. This is caused by the separation of In2O3 nanocrystals from the InSiO3 surface due to the interaction of amorphous In2O3 with water vapor in the air. These In2O3 nanocrystals on the InSiO3 film surface provide clear morphology contrast in both backscattered electron and secondary electron images.

[0046] <2-5-1: Symmetry analysis> Figure 13 shows SEM images of a rotated crystal island whose central region has the (a)

[0222] , (b)

[0400] , and (c)

[0440] crystal orientations, taken with a backscattered electron (BSE) detector, a secondary electron (SE) detector, and the corresponding simulated SEM image (Sim) at an incident electron energy of 15 keV. Miller indices are labeled on the simulated Kikuchi pattern. The lines indicating the edges of each lattice plane in the simulated image show the projection of the lattice plane (hkl) indicated by the Miller indices label. The size of the Brillouin zone (brightest central polygon) in the projection plane of the rotated crystal island with the

[0440] orientation is demonstrated in both the measured SEM image and the simulated image. The characteristic Kikuchi patterns observed in both backscattered electron and secondary electron images clearly originate from Bragg reflections from a series of lattice planes oriented perpendicular to the surface. For the rotated crystal island with a

[0222] orientation shown in Figure 13(a), three major Kikuchi bands emerge from one set of (0-44)-type planes. For the rotated crystal island with a

[0400] orientation shown in Figure 13(b), four major Kikuchi bands emerge from two sets of (004)-type planes and two sets of (0-44)-type planes. For the rotated crystal island with a

[0440] orientation shown in Figure 13(c), six major Kikuchi bands emerge from two sets of (-22-6)-type planes, two sets of (-22-2)-type planes, one set of (-22-2)-type planes, (4-40)-type planes, and (00-4)-type planes.

[0047] Within each island, different sets of Kikuchi bands possess different bandwidths and intersect at "poles," forming bright polygons with filigree-like internal structures. While the shapes of the bright intersection regions of the islands in the

[0222] and

[0400] orientations are quite distinct, the shape of the islands in the

[0440] orientation is relatively complex. The bright intersection region of the islands in the

[0440] orientation is a complex polygon composed of a bright central hexagon, which is the zone axis. This polygon is the intersection region of the (004) Kikuchi band, the (-22-6) Kikuchi band, and the (-22-6) Kikuchi band, as well as six adjacent triangles, which are the intersections of two of these three bands. A series of blurred lines arise from higher-order reflections of the (-440) plane on either side of the main Kikuchi band and can be observed as round bright regions with clear dark contours within the islands in the

[0440] orientation. The intersections between these blurred lines and the major Kikuchi bands form several bright nodes within those major bands.

[0048] Although the contrast of the secondary electron image is significantly lower than that of the backscattered electron image, a highly symmetric Kikuchi pattern with nearly identical features is observed in both the secondary electron image and the backscattered electron image. This result implies that the diffraction information observed in the secondary electron image arises not from the interactions of secondary electrons as they are transported within the crystal, but from cascading secondary electrons generated by backscattered electrons as they traverse the crystal lattice. Therefore, the amount of secondary electrons emitted from different regions of the crystal surface is influenced by the local crystal orientation in which the primary electrons reach different regions of the rotated crystal. The simulated Kikuchi patterns are in good agreement with those experimentally observed in the presented backscattered electron and secondary electron images, showing good agreement between the bandwidths of various types of Kikuchi bands and the angles between them. The simulated patterns also effectively reproduce the fine features of the Kikuchi band intersection regions, i.e., bright complex polygons with filigree internal structures, and round bright areas and their dark contours. Another noteworthy point is that the widths of the major Kikuchi bands observed for rotated crystal islands with the same orientation all remain relatively constant, even far from the center of the edge of the rotated crystal island. The one exception is the (-404) plane of the rotated crystal island with a

[0222] zone axis, due to the lattice distortion, as shown in Figure 13(a). This indicates that the presented rotated crystal island has a very consistent rotation rate, allowing further analysis of the Kikuchi patterns in the SEM images to obtain useful physical quantities related to the crystal structure. Obtaining this information from conventional SEM images of samples is typically impossible.

[0049] <2-5-2: Analysis of lattice constants> According to previous analysis of rotated crystal samples, the Bragg angle can be determined by multiplying the rotation speed of the rotated crystal by the width of the Kikuchi band in the SEM image, as follows:

number

number

[0050] Relational Expression

number

[0222] ,

[0400] , and

[0440] , are 8.49 Å, 13.03 Å ± 2.11 Å, and 11.61 Å ± 2.84 Å, respectively. The average value of the total lattice constant is 11.57 Å ± 2.65 Å. These values ​​are slightly larger than those obtained from the analysis of the bandwidth variation of SEM images of the rotated crystal island at different incident electron energies. [Table 1]

[0051] <2-5-3: Analysis of reciprocal lattice vectors> Based on the determined lattice spacing d, the length of the reciprocal lattice vector |ghkl| can be calculated using the following formula:

number

number

[0052] The conversion factor from the spatial scale to the momentum scale is given by the following relation:

number

[0222] ,

[0400] , and

[0440] , respectively. -1 , 0.053Å -1 , 0.051Å -1 This property allows us to determine the size of the 2D Brillouin zone in the projection plane in terms of momentum scales. Its contour can be determined by extending the prominent Kikuchi bands. Here, the edge lengths of the 2D Brillouin zone in the projection plane of a

[0440] crystal are marked on backscattered electron and secondary electron images (see Fedchenko, O. et al. High-resolution hard-x-ray photoelectron diffraction in a momentum microscope—The model case of graphite. New J. Phys. 21, 113031 (2019)).

[0053] <2-5-4: Strain analysis> A simple estimation of the strain value of the rotated crystal is possible from SEM images of the rotated crystal island by assuming elastic cylindrical bending of the rotated crystal in the radial direction. The maximum strain ε of the surface film can be estimated as follows (see Kolosov, VY & Tholen, AR, Transmission electron microscopy studies of the specific structure of crystals formed by phase transition in iron oxide amorphous films. Acta Mater. 48, 1829-1840 (2000)).

number

[0054] <2-5-5: Analysis of built-in geometrically necessary dislocations (GND)> If we assume that the lattice rotation in a rotated crystal is due to the incorporation of geometrically necessary dislocations (GNDs) that accommodate the lattice curvature caused by the deformation gradient, a lower limit on the GND density can be estimated approximately from SEM images of the rotated crystal (see Nye, JF Some geometrical relations in dislocated crystals, Acta Mater. 1, 153-162 (1953)). Assuming that the GNDs are uniformly distributed within the rotated crystal, a lower limit estimate of the GND density can be obtained by the following formula (see Konijnenberg, PJ, Zaefferer, S., & Raabe, D. Assessment of geometrically necessary dislocation levels derived by 3D EBSD. Acta Mater. 99, 402-414 (2015).):

number

[0055] Therefore, the above equation can be rewritten as follows:

number

number

number

[0222] ,

[0400] , and

[0440] each have a GND density of 1.46x10 15 m -2 , 1.14x10 15m -2 , 1.41x10 15 m -2 Therefore, the total average ground density is 1.34x10 15 m -2 becomes.

[0056] Figure 14 shows detailed information in an SEM image of a rotated crystal island with its central region in the <0400> crystallographic direction, showing (a) the bandwidth, (b) the Bragg angle, (c) the interplanar spacing, (d) the lattice constant, (e) the strain, and (f) the distribution of the built-in geometrically necessary dislocation (GND) density. The Kikuchi patterns in the experimental SEM data exhibit nonuniform bandwidth and shape distortion, as shown in Figure 14, and are slightly different from those calculated by simulation. At the center of the rotating crystal island, the Kikuchi band width and its direction are close to the simulation results. At the edge of the rotating crystal island, the variation in the Kikuchi band width and its direction is significantly larger. This means that in an actual rotating crystal, the rotation speed is nearly the same in the central region, but different at local locations away from the center of the rotating crystal island. Therefore, by measuring the local bandwidth of the Kikuchi band in a specific rotating crystal island, we can determine the distribution of the local Bragg angle, interplanar spacing, lattice constant, strain, and built-in GND density in the region of the rotating crystal island where the Kikuchi pattern appears.

[0057] <3: Cylindrically symmetric rotating crystal and its electron emission direction control> <3-1: Two types of cylindrically symmetric rotation crystals> The first type of cylindrically symmetric rotation crystal 10, as shown in Figure 15A(a), changes direction at each position R, but the azimuthal angle φ remains the same. The magnitude of OR remains the same but changes direction, that is, when R moves along the circumference centered on the reference point O at a constant radius on the crystal film surface, only the azimuthal angle φ changes continuously. The second type of cylindrically symmetric rotation crystal 10 is one in which the [hkl] crystal direction at each position R changes but the polar angle θ remains the same, as shown in Figure 15B(a). As shown in Figures 15A(a) and 15B(a), the crystal structure with the above rotation law is an ideal cylindrically symmetric rotation crystal. In contrast, cylindrically symmetric rotation crystals that are actually manufactured have some distortion relative to the ideal form shown in Figures 15A(a) and 15B(a), as shown in Figures 15A(b) and 15B(b). In other words, forms manufactured so as to basically match the characteristics of the first or second type of cylindrically symmetric rotation crystal 10 also belong to the cylindrically symmetric rotation crystal of the present invention. For example, in the first type of cylindrically symmetric rotation crystal 10, in an ideal cylindrically symmetric rotation crystal, the polar angle θ changes continuously, and the locus of position R where the azimuth angle φ does not change is a straight line. However, in actual production, although the overall shape extends in a straight line, local variations are also allowed, so the locus curves as shown in Figure 15A(b). The region where each azimuth angle φ falls within a certain range is not completely symmetrical. The region where each azimuth angle φ falls within a certain range is, for example, a region within a range of ±5 to 15° of the azimuth angle φ.

[0058] In the second type of cylindrically symmetric rotation crystal 10, in an ideal cylindrically symmetric rotation crystal, the locus of position R when the azimuthal angle φ changes continuously and the polar angle θ does not change is a circle. In actual production, the locus is not a perfect circle, and as shown in Figure 15B(b), the region where each polar angle θ is within a certain range is not a perfect circle but rather an irregular shape based on the circle. The region where each polar angle θ is within a certain range is, for example, a region in the range of ±1 to 5° of the polar angle θ. Possible deformations of the irregular shape based on the circle include stretching, twisting, and local deformation. Furthermore, in an ideal model, the polar angle θ and the azimuthal angle φ are required to change continuously, but in actual fabrication, although there are some discontinuous changes (hopping), the changes tend to be continuous overall, and these crystals can also be considered as cylindrically symmetric rotational crystals.

[0059] <3-2: Differences between cylindrically symmetric rotation crystals and conventional rotation crystals> The cylindrically symmetric rotation crystal of the present invention has a fundamentally different crystal structure from conventionally known rotation crystals. As shown in Figure 16A(b), the cylindrically symmetric rotation crystal 10 used in the present invention does not have the nanofiber regions with clear boundaries that exist in conventionally known rotation crystals, and the overall crystal orientation at adjacent positions rotates continuously. In contrast, as shown in Figure 16B(f), a conventionally known rotated crystal 11 divided into nanofiber regions is actually composed of many nanofiber regions, despite the constantly changing crystal orientation. The crystal orientation within a nanofiber region is nearly uniform, while the crystal orientation changes between nanofibers. This change is discontinuous, and the change in crystal orientation between two nanofibers is irregular [Lutjes, N.R., Zhou, S., Antoja-Lleonart, J., Noheda, B. & Ocelik, V. Spherulitic and rotational crystal growth of quartz thin films. Sci. Rep. 11, 1-12 (2021)].

[0060] 16A and 16B are diagrams illustrating the difference between a cylindrically symmetric rotation crystal and a conventional rotation crystal. In Figure 16A, (a) is an overall perspective view showing one embodiment of a cylindrically symmetric rotated crystal of the present invention, (b) is a schematic diagram of the rotation of the crystal direction in an ideal rotated crystal island, (c) is a radial crystal misorientation angle profile measured on the rotated crystallization island, and (d) is a circumferential crystal misorientation angle profile measured on the rotated crystallization island. In Figure 16A(b), the direction of the arrow represents the direction of any crystal plane. In Figure 16A(c), the dashed curve indicates a reference line for theoretically calculating the deviation of a point around an ideal circularly symmetric rotation crystal from the origin. This deviation angle α is defined by the following equation:

number

[0061] In Figure 16B, (e) is an overall oblique view of a conventional rotated crystal, (f) is a schematic diagram of the crystal direction rotation in a conventionally known rotated crystal island, (g) is the radial crystal misorientation angle profile measured on the rotated crystallization island in Figure 2 of the above-mentioned reference [Lutjes, NR et al., Sci. Rep. 11, 1-12 (2021)], and (h) is the circumferential crystal misorientation angle profile measured on the rotated crystallization island. The rotated crystal film of the present invention shown in Figure 16A(b) forms a special disk-shaped crystallized region where the local crystal orientation changes because the crystallographic branching effect governs the crystallization process. However, in this disk-shaped crystallized region, the local crystal orientation relative to the crystal center direction not only rotates in the radial crystal growth direction, but also permanently rotates around the axis located at the crystal center at the same rotation speed. In contrast, the non-crystallographic branching effect shown in Figure 16B(f) dominates the crystallization growth process in previously known rotated crystal films, forming crystallized regions composed of fibers. The local crystal orientation gradually changes with fiber growth, rotating continuously in the radial growth direction. Due to differences in crystal orientation between fibers, the local crystal orientation does not exhibit consistent rotational behavior in the tangential direction.

[0062] Figures 16A(c) and 16B(g) show the radial misorientation of the rotated crystal. The crystal misorientation curves for these two rotated crystal types exhibit similar apparent linear behavior. From Figures 16A(c) and 16B(g), the gradient of the crystal rotation angle (crystal rotation rate along the selected radius) can be obtained by linear fitting. The crystal orientation rotation rate is 17.6 degrees / μm for the cylindrically symmetric rotated crystal illustrating one embodiment of the present invention, and 0.74 degrees / μm for the rotated crystal in the aforementioned reference [Lutjes, NR et al., Sci. Rep. 11, 1-12 (2021)]. This large difference results from the significant difference in crystal size. The diameter of the cylindrically symmetric rotated crystal island illustrating one embodiment of the present invention is approximately 1.6 μm, while the diameter of the rotated crystal island in the aforementioned reference [Lutjes, NR et al., Sci. Rep. 11, 1-12 (2021)] is approximately 70 μm.

[0063] Figures 16A(d) and 16B(h) show the circumferential misorientation of a rotated crystal. The point-to-point misorientation curves of the cylindrically symmetric rotated crystal island representing one embodiment of the present invention and the rotated crystal in the aforementioned reference [Lutjes, NR et al., Sci. Rep. 11, 1-12 (2021)] are completely different. In the case of the cylindrically symmetric rotated crystal island representing one embodiment of the present invention, the point-to-point misorientation curve is a horizontal line that fluctuates slightly near 1°, with its maximum fluctuation being 2°. The overall accumulated misorientation of continuous rotation vectors about the Z axis at the origin O is 360° per circle. Therefore, the average value of 1° for the point-to-point misorientation curve in the range of 0 to 360° indicates that the observed misorientation is contributed exclusively by local crystal orientation rotation at a nearly constant rate. In other words, the entire rotated crystal island in the example is composed of a single primary fiber growth from a nucleation center, where the crystal orientation rotates without forming small-angle grain boundaries. In the case of the rotated crystals mentioned in the reference [Lutjes, NR et al., Sci. Rep. 11, 1-12 (2021)], the misorientation curves between points fluctuate dramatically, with an average value of 3.42° and a maximum misorientation of 22°. This result indicates that previously known rotated crystals consist of fibers growing radially from a single nucleation point. Therefore, there is no continuous rotation of the crystal orientation along the white circle; only local rotation of the crystal orientation occurs at various speeds within each fiber.

[0064] The dramatic rises and falls observed in the point-to-point misorientation curves are caused by intense, random misorientation fluctuations between adjacent fibers. The point-to-origin misorientation curves for the present rotated crystal are shown in Figure 16A(c) and (d), illustrating how the local crystal orientation rotates along the circumferential direction. The point-to-origin misorientation formed by a vector continuously rotating around the Z axis at origin O is also plotted as a dashed reference curve in Figure 16A(c) and (d), and can be obtained from the following equation:

number

[0065] <3-3: Controlling the electron emission direction of cylindrically symmetric rotating crystals> The charged particle focusing device of the present invention is designed to control the direction of electron emission using a cylindrically symmetric rotating crystal. Experimental measurements and theoretical simulations have shown that the radiation angle of electrons diffracted from within the crystal is primarily affected by the distribution of crystal orientation in the sample. In the case of a typical crystal, the crystal orientation is the same at any position due to translational symmetry, and changing the crystal orientation requires rotating the sample itself. In the case of cylindrically symmetric rotated crystals, the crystal orientation varies at each position, eliminating the need to rotate the sample itself to change the electron diffraction angle. During the crystal growth process, the crystal orientation distribution within each grain and between different grains can be controlled to construct a rotated crystal array, thereby controlling the direction of the electron beam at each position. Furthermore, because cylindrically symmetric rotated crystals are based on Kikuchi diffraction, the shape of this diffraction pattern is not significantly affected by the energy of the incident electrons; that energy only affects the width of the Kikuchi band, not the main direction of the diffracted beam. Therefore, a diffraction element that is not affected by the incident electron energy can be constructed using cylindrically symmetric rotated crystals. In particular, electron diffraction lenses constructed with cylindrically symmetric rotated crystals have low aberrations. When electrons enter a single-crystal material, the incident electron beam interacts with the sample. Some of the electrons undergo inelastic scattering (the electrons change direction with a small loss of energy), resulting in partial waves appearing within the crystal and propagating in all directions in space. If the Bragg diffraction condition is satisfied, these partial waves may also be diffracted at the crystal surface. The larger the scattering angle, the smaller the intensity of the scattered electrons.

[0066] <3-4: Electron beam focusing method for cylindrically symmetric crystals of revolution> Figure 17A is a schematic diagram of the Kikuchi bands formed in a single-crystal thin film material by scattering of an electron beam. (a) shows the relationship between the incident angle of the electron beam (Z beam ) and the vertical axis of the single-crystal thin film material coincide, (b) is the incident angle of the electron beam (Z beam ) and the vertical axis of the single crystal thin film material is tilted. As shown in Figure 17A(a), some of the electrons scattered in all directions in space at point O of the cylindrically symmetric rotational crystal film 10 satisfy the Bragg diffraction conditions for the (hkl) and (-hkl) crystal planes. These diffracted electrons form two conical surfaces 21 (diffraction cones) with their normals as their axes. The conical surfaces 21 are blocked by the fluorescent screen 20, and their intersection lines become two hyperbolas. Because the sample is very far from the fluorescent screen 20, these two hyperbolas 21 are approximated as straight lines, forming Kikuchi bands.

[0067] Intersections of Kikuchi bands produce more intense scattering spots, which vary depending on the angle between the electron beam and the crystal plane. Thus, as shown in Figure 17A(b), if a specific position x on the film 10 follows a specific relationship with the local crystal plane direction θ, the intense scattering spots generated by the electron beam crossing this position x will coincide at a specific distance f. Placing a fluorescent screen 20 at this specific distance f will produce a diffraction spot image. In this way, the electron beam is focused to a focal length f.

[0068] FIG. 17B(c) is a schematic diagram of an ideal polycrystalline material capable of focusing an electron beam, and FIG. 17B(d) is a schematic diagram of a rotating crystal capable of focusing an electron beam. Intersections of Kikuchi bands produce more intense scattering spots, which vary with the angle between the electron beam and the crystal plane. Thus, as shown in Figure 17B(c), if a specific position x on film 10 follows a specific relationship with the local crystal plane direction θ, the intense scattering spots generated by the electron beam crossing this position x will coincide at a specific distance f. Placing a fluorescent screen 20 at this specific distance f will produce diffraction spot images 20a, 20b, and 20c.

[0069] That is, as shown in FIG. 17B(c), the electron beam is incident perpendicularly to the film material (Z beam ), scattering occurs in three cylindrically symmetric rotated crystal regions 10a, 10b, and 10c of film 10, forming a daisy cell pattern. If the positions of the three cylindrically symmetric rotated crystal regions are O, x1, and x2, and the angles between the respective crystal planes and the incident electrons are 0, θ1, and θ2, respectively, the electrons are first scattered at the origin O, and then most strongly scattered in the same direction as the most strongly scattered direction of the incident electrons. When the relative crystal orientations θ1 and θ2 at the x1 and x2 positions on a cylindrically symmetric rotation crystal film satisfy the following equation, the special rotation crystal film can converge an electron beam and can be regarded as a convex lens for the electron beam. x1 / tanθ1=x2 / tanθ2 (17) After the electron beams from positions x1 and x2 pass through this cylindrically symmetric rotational crystal film, at a distance f from the lower surface of the film (ignoring the film thickness), part of the electron beam converges at a distance f below the zero position of the film. f=x1 / tanθ1=x2 / tanθ2 (18) Here, f can be called the focal length of the cylindrically symmetric rotation crystal thin film lens. In other words, if the crystal orientations θ1 and θ2 at positions x1 and x2 on the film satisfy equation (18), when an incident electron scatters at position xi (i = 1, 2), the most strongly scattered electron is shifted in the direction of θi from the incident direction, and three strongly diffracted spots overlap at a distance f from the film.

[0070] The electron beam is focused at f, where f is the focal length of a special thin-film lens. In the case of this cylindrically symmetric rotational crystal film, the distance between x1, x2 and O is much smaller than the focal length f, and therefore the relative orientations θ1, θ2 of the crystals are small, so the focal length f can be approximately calculated by the following equation: f=x1 / θ1=x2 / θ2 (19)

[0071] Cylindrically symmetric rotating crystals do not satisfy translational invariance but do satisfy rotational symmetry. Figure 17B(d) illustrates the principle of electron beam focusing using a specially constructed rotating crystal, showing the Kikuchi patterns of four parallel electron beams passing through a 30 nm thick InSiO crystal 10. These four patterns were obtained using the Monte Carlo method for theoretical simulations of electron scattering trajectories. The theoretical simulation of electron scattering trajectories was based on that disclosed in Phys. Chem. Phys. 17 (2015) 17628. In this special structure of the rotational crystal film, the rotation angle of the crystal plane at a certain local position is θ, and the distance between this local position and the origin O is r. Assume that θ and r in the entire region of the rotational crystal film 10 always satisfy the following equation. r / f=tanθ (f is a constant) (20) Then, when an incident electron beam passes through the rotating crystal films 10d, 10e, 10f, and 10g, the resulting strong scattering points will coincide at the distant positions 20d, 20e, 20f, and 20g, thereby achieving maximum efficiency in electron beam focusing.

[0072] <4: Theoretical simulation method for electron diffraction from cylindrically symmetric crystals of revolution> The present invention proposes a theoretical simulation method that allows the simulation of electron diffraction intensity in cylindrically symmetric crystals of revolution. According to the dynamical method of electron diffraction, the wave function of the electrons inside the crystal can be written in the form of Bloch waves as follows:

number

number

[0073] where:

number

number

number

[0074]

number

number

number

[0075] The probability density P(r) at any position in the crystal is:

number

number

[0076] In general mechanical simulations, the direction of the wave vector K is defined with respect to the ground as the reference frame. Taking parallel beam incidence as an example, due to the translational symmetry of conventional crystals, the angle between the wave vector K and any crystal orientation is constant at any position with parallel beam incidence. However, in cylindrically symmetric rotation crystals, translational symmetry is broken, and the crystal orientation may vary at each position within the sample. This makes it impossible to define the direction of the wave vector K with respect to the ground as the reference frame. Here, we provide a new definition, as shown in Figure 9. When a crystal orientation is selected as the reference, for example, the crystal orientation [hkl] at the symmetry center position O is used as the reference, the direction of [hkl] is the direction of K at any position within the sample relative to the incident wave vector K. For conventional crystals, this definition is the same as the direction defined with respect to the ground as the reference frame. Therefore, this method is effective not only for simulations of conventional crystals but also for special crystals where the crystal orientation changes depending on the location. This method allows the structure and properties of materials to be designed and evaluated through computer simulation, providing effective guidance for product design and development and reducing input costs in actual production.

[0077] <5: Method for producing cylindrically symmetric rotation crystals> The method for producing a rotated crystal of the present invention can control the orientation of the crystal during crystal growth and produce a cylindrically symmetric rotated crystal. Figure 18 is a flow chart showing the growth process of a rotated crystal, illustrating the deposition process of an amorphous thin film of InSiO by magnetron sputtering. A 30-nm-thick amorphous film was fabricated by DC magnetron sputtering (Shibaura Mechatronics, CFS-4EP-LLi-Miller) on a sapphire substrate 14 at room temperature. Sputtering targets 17, 18 consisting of In2O3 and SiO2 were used. The Si / In ratio in the sputtering target was 2.3 at.%, which corresponds to 1 mass%. Here, % is calculated as SiO2 / (In2O3 + SiO2).

[0078] In the sputtering apparatus, the sputtering targets 17 and 18 were spaced 160 mm apart from the substrate 14. The InSiO film 15a was produced by plasma at 200 W in an argon / oxygen mixed atmosphere with a gas flow ratio of 1:1 and a total pressure of 0.25 Pa. This gas flow ratio was determined to produce an electrically stable InSiO film 15a that was resistant to thermal stress. In situ SEM observation was performed to obtain images of dynamic crystallization along a precisely fixed observation area. The InSiO film was heated in situ to 300 °C in the SEM environment. This allowed us to control the crystal orientation during InSiO crystal growth, resulting in cylindrically symmetric rotated crystals, as shown in Figures 24 and 25A.

[0079] <6: Device for electronically manipulating thin films of cylindrically symmetric rotation crystals> The present invention proposes an electron manipulation device made of a cylindrically symmetric rotated crystal. Using the above-described fabrication method, the crystal orientation distribution at each position changes during crystal growth in a cylindrically symmetric rotated crystal 10, as shown in Figure 19. Electrons incident from different positions diffract in different directions. Because the angular distribution of diffracted electrons differs depending on the crystal orientation distribution, electrons incident at each position can be controlled and manipulated. Figure 19a shows a schematic diagram of a device in which electrons are manipulated by a single crystal island of a cylindrically symmetric rotated crystal 10, while Figure 19b shows a schematic diagram of a device in which electrons are manipulated by an array of crystal islands 10h, 10i, and 10j.

[0080] <7: Method for growing cylindrically symmetric rotating crystal focusing lenses in an array> The present invention proposes a method for growing aligned rotating crystal electron focusing lenses to obtain a compact electron beam splitting system. Figure 20 is a flow chart showing the manufacturing process of a rotated crystal film lens array. If a single rotated crystal is called a rotated crystal single crystal, a rotated crystal array can also be called a rotated crystal polycrystal. A rotated crystal array can be grown by sputtering. The specific growth method is to deposit an amorphous thin film of rotated crystal by magnetron sputtering. When heating the amorphous rotated crystal thin film, the heating temperature is controlled to a temperature slightly lower than the crystallization temperature T0 (350°C), and a high-intensity electron beam is irradiated onto the amorphous thin film according to the designed array period. When the local temperature of the electron beam irradiation reaches the crystallization temperature, the crystals diffuse outward from the irradiation point, forming a rotating crystal.

[0081] The heating temperature by electron beam irradiation was controlled to be slightly lower than the crystallization temperature of 350° C., and a high-intensity electron beam 16 was irradiated onto the amorphous film 15 in accordance with the designed array period. The method for growing a rotating crystal electron focusing lens is as follows: Rotating crystals are arranged and multiple electron beams 16 are focused onto them. An SEM equipped with a heated sample stage system is used to grow an array of rotating crystal films 10 in specific areas by irradiating and heating with the electron beams. Magnetron sputtering is used to deposit an InSiO amorphous film 15 on a substrate 14. 20, when heating an InSiO amorphous film 15, the heating temperature is controlled to be slightly lower than the crystallization temperature T0, and a high-intensity electron beam 16 is irradiated onto this amorphous film 15 according to the designed arrangement period. When the local temperature at the collision point of the electron beam 16 reaches the crystallization temperature, the crystals spread outward, forming a crystal that rotates around the collision point.

[0082] <8: High-throughput preparation and detection method using cylindrically symmetric rotation crystals> This invention provides a high-throughput method for producing thin films of cylindrically symmetric rotation crystals using a magnetron sputtering system. The inventors have successfully grown a rotation crystal film with a very unique crystal structure (a structure in which the crystal plane rotates depending on the position on the crystal film) using a beam exposure method with PMMA, and confirmed the ability to focus an electron beam using simple electrons. In this invention, the inventors have diligently conceived of further improving growth parameters, optimizing focusing efficiency, developing a measurement device capable of quantitatively measuring focusing efficiency, and realizing and applying the growth of a rotation crystal film lens array using an electron beam and heating method. This conceived method is applied to electron beam lithography.

[0083] Figure 21 is a flow chart showing the growth process of a rotational crystal film lens, with each step shown in Figures 21A to 21E. As shown in Figure 21A, it is fabricated using a magnetron sputtering dual-target method. Specifically, an InSiO amorphous film 15 is deposited on a substrate 14 using an InO2 target 17 and an SiO2 target 18. The Si content of the dopant element in the InSiO amorphous film varies continuously with position in the film. When heating the amorphous InSiO film, it is first heated to a temperature slightly lower than the crystallization temperature T0, and then a high-intensity electron beam 16 is irradiated onto the amorphous film 15 according to the designed array period (Figure 21B). When the local temperature of the electron beam reaches the crystallization temperature, the crystals spread outward around the irradiation point, forming a rotating crystal. By varying the intensity of the incident electron beam during the electron beam bombardment heating process, rotating crystal lenses of InSiO amorphous film 15 can be fabricated at different heating temperatures. Because the Si content of the doping element in the InSiO amorphous film 15 varies depending on the local location within the film, multiple rotating crystal lenses can be fabricated in a single InSiO rotating crystal film fabrication, forming an array of rotating crystal lenses with each lens having a different Si impurity content and heating temperature.

[0084] First, a rotated crystal InSiO film 10 grown on a substrate was transferred to a TEM grid 19 (Figure 21C). Theoretical calculations suggest that a 1 μm diameter rotating particle can focus a 30 kV electron beam to a spot with a half-width of 16 nm, located approximately 5 μm away from the thin film material. The ultra-short focus (micron level) and ultra-fine focus size of 16 nm make direct observation difficult using existing 2D detectors (with minimum pixel sizes on the micron order). To accurately verify the experimental transmission efficiency and minimum size of this focusing lens, we modified the SEM sample stage as shown in Figures 21D and 21E. The SEM sample stage 30 includes a semiconductor detector 31 (SSD: Solid State Detector), an aperture 32, a silicon thin film 33, a cylindrically symmetric rotational crystal layer 34, a TEM mesh 35, a TEM mesh support 36, and a silicon thin film support 37. An electron beam 38 is irradiated onto the cylindrically symmetric rotational crystal layer 34 via the TEM mesh 35 and focused onto the semiconductor detector 31 via the aperture 32. The silicon thin film 33 is supported by the silicon thin film support 37, and its position and orientation on the SEM sample stage 30 are maintained. The TEM mesh 35 is supported by the TEM mesh support 36, and its position and orientation on the SEM sample stage 30 are maintained. The cylindrically symmetric rotational crystal layer 34 is supported by the TEM mesh support 36 via the TEM mesh 35. This detection system offers higher contrast than conventional methods that use a secondary electron detector in the SEM body to detect backscattered electrons on a silicon thin film, and Monte Carlo simulations have shown that the maximum spot size is less than 10 nm.

[0085] <9: X-ray and fluorescence light source device using cylindrically symmetric rotation crystal film> The present invention proposes an apparatus for focusing an electron beam by a cylindrically symmetric rotational crystal film in order to obtain a more uniform light source. Figure 22 is a diagram illustrating the relationship between the size of the focus of an electron beam and the degree of blur in an image. In the figure, an electron microscope 40a is equipped with an electron gun 41, an anode 42, a target 43, a substrate 44, and a detector 46, and images an object 45 with an electron beam. The target 43 is stacked on the substrate 44. Some microscopes are equipped with an electromagnetic lens 47 for focusing on the object 45 (Figure 22B). In an X-ray tube, the size of the electron beam focus affects the degree of image blur (Figure 22A). Generally, in X-ray tube transillumination, the smaller the focus size, the higher the resolution. Increasing the resolution allows for more detailed images to be obtained. To reduce the electron beam focus size, a microfocus X-ray tube with the structure shown was developed for the electron microscope 40b shown in Figure 22B. The most important component of a microfocus X-ray tube is the electron optical circuit, and although the electron gun 41 has been successfully miniaturized, the centimeter size and focal length of the electron lens system 47 (electromagnetic and electrostatic lenses) pose an obstacle to miniaturizing the entire microfocus X-ray tube system.

[0086] The electron microscope 40c shown in Figure 22C illustrates a microfocus X-ray tube with a structure embodying one embodiment of the present invention. This microfocus X-ray tube utilizes the electron diffraction lens, which utilizes a unique property: when an amorphous nanofilm is heated by an electron beam, a special crystallization occurs. The area of ​​the amorphous nanofilm irradiated with the electron beam spontaneously forms an electron diffraction lens that focuses the electron beam. This property allows the electron diffraction lens 48, an embodiment of the present invention, to be applied to a metal target 43 as a coating material. The heat from the incident electron beam induces special crystallization in the coating material, spontaneously forming a microelectronic optical circuit that focuses the incident electron beam. This process does not require an optical circuit calibration process or a micro-nanofabrication process to reduce the electron beam focus size within the X-ray tube, allowing for the realization of a microfocus X-ray tube at very low cost.

[0087] Electron beam focusing has always been a highly valuable technology. Nearly all existing devices that use electron beams as detection signals require electron beam focusing techniques to focus the emitted or collected electron beams, thereby improving the efficiency of these devices. Because electrons have a negative charge, they can be very easily focused using an electromagnetic field. Currently, the most common electron beam focusing techniques are electromagnetic and electrostatic lenses. When an electron beam is incident parallel to a thin-film material, the electrons are scattered within the material, deflecting their direction of travel, causing the beam to diverge after passing through the material. In other words, any thin-film material can act as a concave lens for electrons. Therefore, spherical aberration in electron microscope images can be corrected by adjusting the thickness of the thin-film material to a certain extent. Previous academic literature in this field has used thin film materials as concave lenses, which can only scatter light on the electron beam and therefore only compensate for the focusing effect of conventional electromagnetic or electrostatic lenses. This patent specification describes a convex lens for electrons that can focus the electron beam and therefore can independently configure the electron optical path without relying on conventional electromagnetic or electrostatic lenses. [Example]

[0088] <10: Example 1: InSiO Cylindrically Symmetric Rotation Crystal> Amorphous films with a thickness of 30 nm were fabricated on sapphire substrates by DC magnetron sputtering (Shibaura Mechatronics, CFS-4EP-LLi-Miller) at room temperature. A sputtering target consisting of In2O3 and SiO2 was used. The Si / In ratio in the sputtering target was 2.3 at.%, which corresponds to 1 wt.% SiO2 / (In2O3 + SiO2). The sputtering target and the substrate were spaced 160 mm apart in the sputtering system. InSiO films were produced by plasma deposition at 200 W in an argon / oxygen atmosphere with a gas flow ratio of 1:1 and a total pressure of 0.25 Pa. This gas flow ratio was determined to produce electrically stable InSiO films resistant to thermal stress. Finally, the InSiO films were annealed at 300 °C to produce InSiO films with a cylindrically symmetric crystal structure. The annealing process was simultaneously performed using in situ SEM, allowing dynamic crystallographic observation.

[0089] Figures 20A(b) and 20B(b) show ESBD relative crystal misorientation maps of the cylindrically symmetric InSiO crystal island prepared by the above method. It is clearly observed that the crystal plane rotates continuously in both the polar angle θ and the azimuthal angle φ directions with respect to the center. Figure 23 is an explanatory diagram of a rotated crystal island. Figures 23(a) and (b) show that in the TEM images of the cross section of the rotated crystal, the angle between the (400) plane (shown by the hollow broken line) of the rotated crystal (InSiO) cross section and the (100) plane (shown by the hollow long broken line) of the sapphire cross section changes depending on the observation position of the TEM image. A fast Fourier transform (FFT) was performed on these cross section images, and the change in the (400) plane of the rotated crystal cross section is shown in the figure below (Figure 23(c)). Figure 23(a) shows a cross section of a rotated crystalline island in the <0400> direction, with the central region being a rotated crystalline island and the adjacent side regions being amorphous regions. Regions I to V of the cross-sectional profile of the rotated crystalline island in the <0400> direction in Figure 23(a) were enlarged and observed using a TEM. A magnified TEM image is shown in Figure 23(b). In Figure 23(b), the lattice structure of the sapphire substrate can be clearly seen along with the InSiO crystalline regions in regions I to V. In the TEM cross-sectional image in Figure 23(b), the angle between the (400) plane of the InSiO cross section (marked by the short white dashed line) and the (100) plane of the sapphire (marked by the long white dashed line) changes depending on the observation position of the TEM image. The lattice constants measured from the TEM images are 0.714 nm for InSiO and 0.435 nm for the sapphire substrate.

[0090] As shown in Figure 23(c), a fast Fourier transform (FFT) was performed on the cross-sectional images of regions I through V to more clearly observe the changes in the (400) plane of the InSiO cross section. The thick black dashed line indicates the

[0400] direction in momentum space after FFT, the long white dashed line indicates the (100) plane of the sapphire substrate, and the thin black dashed line indicates the horizontal reference. The angle between the

[0400] direction in momentum space of the rotated crystal after FFT and the (100) plane of the sapphire substrate is shown in the bottom row of each of regions I through V in Figure 23(c). When the observation position of the TEM image is at the left end of the rotated crystal island, the angle is 97.0°. As the observation position moves to the right, the corresponding angle decreases at a nearly constant rate, reaching 83.1° when the observation position is moved to the right end of the rotated crystal island (680 nm from the left end). When the observation position within the rotating crystal island is changed, the (400) crystal plane within the rotating crystal island rotates continuously in the cross-sectional direction, and the rotation speed is approximately 20.4° / μm. The unique rotational properties of the crystal planes of the cylindrically symmetric rotated crystals described above are evident from EBSD crystal deflection measurements via direct observation of the TEM profile. The specific properties are explained in detail in the previous chapter on rotated crystal concepts. [Example]

[0091] <11: Example 2: Images generated by an electron beam interacting with a cylindrically symmetric crystal of rotation> Cylindrical symmetric rotation crystals interact with an incident electron beam to generate a Kikuchi diffraction pattern. Kikuchi diffraction patterns typically require specific methods, such as grazing incidence methods like electron backscatter diffraction, rocking beam methods like electron channel diffraction, and convergent beam methods like convergent beam electron diffraction. The common thread between these methods is that electrons must diffract from the crystal surface at multiple angles, and the diffracted waves interfere with each other to ultimately form the Kikuchi diffraction pattern. In contrast, cylindrically symmetric rotated crystals offer a completely new approach: instead of changing the angle of the incident electron beam itself, the angle of the crystal surface inside the crystal is changed so that the electrons are diffracted by the crystal surface at many angles inside the crystal. The difference is that while with ordinary crystals the Kikuchi diffraction pattern cannot be observed in the grating scan mode of an SEM, with cylindrically symmetric rotated crystals the Kikuchi diffraction pattern can be obtained in the grating scan mode of an SEM.

[0092] FIG. 24 shows an SEM image of the rotated crystal in the experiment, showing an SEM image of a crystal island of an InSiO cylindrically symmetric rotated crystal obtained with a backscattered electron detector. Figure 24(a) shows the crystallization of a rotated crystal (InSiO) during the annealing process. Films grown on these substrates partially crystallized, with a large amount of material remaining amorphous. When the films were annealed at 300 °C, the amorphous film gradually crystallized and grew into circular crystalline islands approximately 1–2 μm in diameter, forming a distinct Kikuchi pattern. These rotated crystalline islands have a circular structure and grow radially and quasi-anomalously from a single nucleation point. The growth of these rotated crystalline islands is reasonably considered to initiate from nucleation points randomly distributed throughout the amorphous film and grow two-dimensionally from the center. In some cases, nucleation sites are isolated, while in other cases, growth is terminated due to the presence of another crystallographic site from neighboring nucleation sites. Rotated crystal islands crystallized in different regions show different Kikuchi diffraction patterns, indicating different crystal orientations. The indices in parentheses [hkl] represent the crystal orientation at the position indicated by the arrow.

[0093] Whether the experimentally observed Kikuchi pattern is due to electron diffraction can also be verified by observing the movement of the Kikuchi pattern when the sample stage is tilted. Figure 24(b) shows SEM images of a rotating crystal island in the central region near the axis of the

[0440] region when the sample stage is tilted from -5° to 5°. Focusing on the brightest central complex polygon (the intersection region of multiple Kikuchi bands) in the SEM image, we can see that the brightest central polygon moves with the tilt of the sample stage. Taking 5° as the origin, we observe that the brightest central polygon moves downward with each change in the sample stage angle.

[0094] To further investigate the electron energy dependence of the observed Kikuchi patterns, SEM images of the rotated crystal island in the central region near the

[0440] zone axis at different incident electron voltages are shown in Figure 24(c). It can be seen that the widths of all Kikuchi bands observed in these SEM images gradually narrow as the incident electron energy increases in the range from 5 keV to 30 keV, and the size of the brightest central hexagonal region formed by the intersection of these Kikuchi bands decreases. This is in accordance with Bragg's theorem:

number

[0095] The different orientations of the Kikuchi diffraction patterns for each crystal island and their variations with sample angle and incident electron energy can also be simulated using the theoretical methods described herein. Figure 25 shows the simulated Kikuchi diffraction patterns for an ideal InSiO2 cylindrically symmetric crystal. Figure 25(a) shows an SEM image of a rotated crystal theoretically simulated to represent the diffraction patterns of each crystal island in Figure 24(a). To construct the SEM pattern of the rotated crystal, we simulated the diffraction process of electrons incident on an In2O3 single crystal with different directions to obtain an electron channeling pattern (ECP). In this simulation, fluctuations in the crystal rotation speed of each crystal island were ignored, and the average rotation speed, i.e., the uniform change in the angle between the electron incident direction and the In2O3 single crystal lattice, was used. Details of the simulation method are already described in the publicly known publication "Bragg WL The Diffraction of Short Electromagnetic Waves by a Crystal. Proc. Camb. Philos. Soc. 17, 43-57 (1913)." The reconstructed simulated images closely reproduced the experimental SEM images, and the number and distribution of Kikuchi bands in each island were consistent with those in the reconstructed images, suggesting that the Kikuchi patterns in these SEM images were indeed induced by crystal rotation. On the other hand, the Kikuchi patterns in the experimental SEM images exhibited nonuniform band widths and distorted shapes, differing from those in the simulated images. While the bandwidth and orientation of the Kikuchi bands in the center of the rotated island were close to those in the simulated images, the variations in bandwidth and orientation of the Kikuchi bands at the edges of the island were clearly large. This indicates that the rotation speed of an actual rotated crystal is nearly constant in the central region but differs at local locations away from the center of the island.

[0096] Figure 25(b) shows the simulated ECP results for an In2O3 single crystal at corresponding tilt angles of the sample stage along the

[0440] zone axis, corresponding to the simulated Kikuchi pattern displacement for different tilt angles of the sample in Figure 24(b). The simulated Kikuchi pattern along the

[0440] zone axis (spanning a 30° angle) of In2O3 is similar to the experimental SEM pattern in terms of Kikuchi band bandwidth and the size of the brightest central polygon region. This suggests that the local crystal orientation of this rotated crystal island rotates by approximately 30° from one side to the other. Furthermore, the movement of the brightest central polygon in the simulated Kikuchi pattern is downward when the tilt angle of the sample stage is changed. The movement distance of the brightest central polygon with varying the tilt angle of the sample stage is essentially the same as that observed experimentally.

[0097] Figure 25(c) shows the ECP simulation results for the axial direction of the

[0440] zone of an In2O3 single crystal at different incident electron energies, corresponding to the simulation of the Kikuchi band width for different incident electron energies in Figure 24(c). It can be seen that the Kikuchi band width and the size of the brightest central polygonal region in the simulated Kikuchi patterns obtained at different incident electron beam energies are essentially consistent with the experimentally observed Kikuchi pattern. This indicates that the rotated crystal island with a diameter of approximately 1.64 μm shown in the SEM image has a rotation span of approximately 30 degrees along the diameter direction under the pattern.

[0098] Figure 25(d) shows the experimental and simulation results of the relationship between the tilt angle of the sample and the displacement of the Kikuchi pattern. Figure 25(e) shows the experimental and simulation results for the width of a certain Kikuchi band at different incident electron energies. We can see that the simulation results are in excellent agreement with the experimental results, demonstrating that the simulation method described herein can effectively predict and analyze the diffraction patterns of cylindrically symmetric rotation crystals. [Example]

[0099] <12: Example 3: Cylindrically symmetric crystal of revolution for producing an electron diffraction lens with a single crystal island> Cylindrically symmetric rotation crystals have different crystal orientations at each position, and can be used for various purposes if the crystal orientations are distributed in various ways. In the process of crystal growth, the crystal orientation of local crystallization is controlled as shown in Figure 26. If the focal length is f, then the distance between the sample surface and the reference point O of the symmetry center is:

number

[0100] Taking a bulk electron gun as an example, a bulk electron gun is an electron source that emits electrons in all directions. The electrons incident on each position on the sample are approximated as spherical waves, and the emitted electrons are emitted in all directions. The electron diffraction intensity distribution of the bulk electron gun, which is the electron source after passing through the electron diffraction lens of this embodiment, was simulated using the theoretical simulation method for electron diffraction from a rotating crystal proposed in this patent. Figure 26 shows the simulation results of the exit intensity cross section in the xy direction (propagation direction is the z direction) of an electron beam emitted from the bulk electron gun after entering the electron diffraction lens. Figure 26 shows the distance from the propagation direction z, from a bulk electron gun with electron energy of 15 keV, after passing through an electron diffraction lens consisting of crystal islands of a 30 nm-thick InSiO film, from 0.2, 0.4, ..., to 1.8f, with focal length f as the reference. The emitted electron beam gradually exhibits a constant distribution as the propagation distance increases. When it propagates to focal length f, a bright spot appears at the center of the cross section, and a clear Kikuchi diffraction pattern appears. As the propagation distance continues to increase, the intensity of the electron beam becomes dispersed and the Kikuchi diffraction pattern becomes obscured as it is already overfocused. The bright spots in this Kikuchi diffraction pattern are located on the target surface layer and are the source of the x-ray radiation.

[0101] Figure 27 shows the change in electron intensity on the central axis in the propagation direction (z direction) after the output intensity of the bulk electron gun enters the electron diffraction lens. Immediately after the sample is released, that is, when z is small, the intensity on the central axis is very low. As the propagation distance increases, the intensity on the central axis gradually increases. As can be seen from the peak in Figure 27, the intensity of the output electrons reaches a peak when they propagate to the focal length, that is, when they propagate to the focal position. When passing through the focal point, the intensity of the electron beam is dispersed, and the intensity of the output electrons on the central axis decreases. From Figures 26 and 27, it is clear that the electron diffraction lens with a single crystal island made from a cylindrically symmetric rotation crystal has an excellent focusing effect on the electron beam and does not require any special requirements for the electron source. [Example]

[0102] <13: Example 4: Cylindrically symmetric rotating crystal for producing a plurality of rotating crystal island arrays for electron diffraction lenses> During the crystal growth process, the crystal orientation of local crystallization is controlled as shown in Figure 28. If the focal length is f, then the distance between the sample surface and the reference point O of the symmetry center is:

number

[0103] <14: Example 5: Cylindrically symmetric rotation crystal for fabricating electron beam splitter> During crystal growth, the crystal grows around multiple symmetry centers. As an example, three symmetry centers O1, O2, and O3 are selected for growth, as shown in Figure 29. For any position on the crystal surface, the distance x between that position and the nearest symmetry center is defined as

number

[0104] The cylindrically symmetric rotation crystal used in the electron beam focusing device of the present invention can be used as a cylindrically symmetric rotation crystal island array electron diffraction lens as an electron beam focusing device. The electron beam focusing device can effectively increase the total area of ​​the focusing crystal islands, and the larger the total area of ​​the crystal islands, the stronger the electron focusing strength. At the same time, the cylindrically symmetric rotation crystal island array electron diffraction lens is more flexible in terms of spatial arrangement and does not need to occupy the entire area. In addition, the remaining space can be used as a window for other signal sources (e.g., ion beams, lasers) or detector signals, allowing for flexible implementation of various experimental techniques. The microfocus X-ray tube of the present invention is configured such that an incident electron beam is converted into X-rays in the region where bright spots of the Kikuchi diffraction pattern formed on the surface layer of a metal target by a cylindrically symmetric rotation crystal are generated, and the X-rays are projected onto the subject as a roughly X-ray point source. This allows for a structure that projects X-rays onto the subject at a focal length that is significantly shorter than that of conventional electromagnetic lenses, which contributes to the miniaturization of electron microscopes, for example. [Explanation of symbols]

[0105] 10 Cylindrically symmetric rotating crystal 10h, 10i, 10j Crystal island array 11 Rotated crystals divided into nanofiber regions 14 PCB 15 InSiP layer 16 Electron Beam 20 fluorescent screen 20a, 20b, 20c Diffraction spot images 21 Hyperbola 30 SEM stage 31 Solid State Detector (SSD) 32 Aperture 33 Silicon thin film 34 Cylindrically symmetric rotating crystal layer 35 TEM mesh 36 TEM mesh support 37 Silicon membrane support 38 Electron Beam 40a, 40b, 40c Electron microscope 41 Electron Gun 42 Anode 43 Target 44 PCB 45 Object 46 detectors 47 Electromagnetic Lens 48 Cylindrically symmetric rotating crystal layer

Claims

1. A rotation crystal film having a cylindrically symmetric rotation crystal with a thin plate-like planar structure cut from the center (O) of a spherical crystal along a plane direction perpendicular to the north pole direction (Z), the thickness of the thin plate-like planar structure being 5 nm or more and 1 mm or less; An electron beam focusing device characterized in that an incident electron beam passes through the cylindrically symmetric rotation crystal, and the resulting strong scattering point coincides at a focal position (f) that is farther away than the film thickness of the cylindrically symmetric rotation crystal.

2. An arbitrary point (R) of the thin plate-like structure is a predetermined distance (r) away from the center (O) of the spherical crystal, is inclined by a polar angle (θ) with respect to the north pole direction (Z), and is expressed as a position inclined by an azimuthal angle (φ) with respect to the meridian connecting the north pole and south pole directions (Z) of the spherical crystal. The incident direction of the incident electron beam (Z ｂｅａｍ ) and the polar angle (θ) and a predetermined distance (r) of the rotational crystal film with respect to the out-of-plane normal direction [hkl] of the rotational crystal film approximately satisfy the following formula: r / f = tan θ (f is a constant) 2. The electron beam focusing device of claim 1.

3. The crystal orientation [hkl] of the cylindrically symmetric rotation crystal coincides with at least one of [111], [110], [100], [211], [121], and [310] when the incident direction of the incident electron beam is [001].

3. The electron beam focusing device of claim 2.

4. The cylindrically symmetric rotation crystal does not have a nanofiber region having a crystal boundary, and the overall crystal orientation at adjacent positions in the azimuthal angle φ direction is continuously rotated starting from a reference point O.

2. The electron beam focusing device of claim 1.

5. The locus of positions of radius R where the polar angle θ changes continuously and the azimuthal angle φ does not change is linear.

3. The electron beam focusing device of claim 2.

6. When the azimuth angle φ changes continuously and the polar angle θ does not change, the locus of the position of R is approximately a circle.

3. The electron beam focusing device of claim 2.

7. The thickness of the cylindrically symmetric rotation crystal is 5 nm or more and 1 mm or less, The multilayer material of the rotational crystal film consists of Si-doped InO and unavoidable impurities.

2. The electron beam focusing device of claim 1.

8. The rotation crystal film has a cylindrically symmetric rotation crystal that is covered with an amorphous film, and the amorphous film has another cylindrically symmetric rotation crystal that is different from the cylindrically symmetric rotation crystal.

2. The electron beam focusing device of claim 1.

9. 10. A microfocus X-ray tube comprising an electron gun and a metal target coated with the rotational crystal film of claim 1 to a predetermined thickness, The cylindrically symmetric rotation crystal coated on the metal target has a diameter of 1 μm or more and 20 μm or less, an incident electron beam emitted from the electron gun is incident on the cylindrically symmetric rotation crystal, and the incident electron beam is converted into X-rays in a region where bright spots of a Kikuchi diffraction pattern are generated on a surface layer of the metal target by the cylindrically symmetric rotation crystal, and the X-rays are projected onto an object as a substantially X-ray point source; Microfocus X-ray tube.

10. The surface layer of the metal target is The polar angle (θ) of the rotation crystal film and the predetermined distance (r) are a depth region from the surface including a focal length f that approximately satisfies the following formula: r / f = tan θ (f is the focal length) 10. The microfocus x-ray tube of claim 9.

11. The crystal orientation [hkl] of the cylindrically symmetric rotation crystal coincides with at least one of [111], [110], [100], [211], [121], and [310] when the incident direction of the incident electron beam is [001].

10. The microfocus x-ray tube of claim 9.

12. The cylindrically symmetric rotation crystal is surrounded by an amorphous film, The cylindrically symmetric rotation crystals are crystallized in different regions of the amorphous film, and the cylindrically symmetric rotation crystals crystallized in the different regions each exhibit different crystal orientations, The Kikuchi diffraction patterns are present at positions (r, φ, θ) corresponding to the respective crystal orientations of the cylindrically symmetric rotation crystal.

10. The microfocus x-ray tube of claim 9.

13. a focal distance from the rotational crystal film to which the incident electron beam is focused (the distance between the interface between the thin film material and the target material and the position where the electron beam is most focused) of 50 nm or more and 1 mm or less; The focal spot size of the electron beam in the X-ray tube is 1 nm or more and 100 μm or less.

10. The microfocus x-ray tube of claim 9.

14. 14. The microfocus X-ray tube according to claim 13, wherein the metal target is at least one of Cr, Fe, Co, Ni, Cu, Mo, Ag, and W, or a material that may become an X-ray target material in the future.

15. 10. The fine focus x-ray tube of claim 9, wherein the electron diffraction lens operates as a microelectronic optical circuit to focus an incident electron beam.

Images