Beam Electron Diffraction Patterns Research Articles

Performance of a deep convolutional neural network to classify crystal structures using selected area electron beam diffraction patterns containing lattice defect information.

A deep convolutional neural network (DCNN) architecture ResNet has been tested to verify its ability to handle selected area electron diffraction pattern (SADP) datasets carrying information on lattice defects including strains, thermal lattice vibrations, point defects, dislocations, and twin boundaries. The disordered states of the crystal lattices in the presence of these defects were predicted by ab initio molecular dynamics simulations, first principles geometry optimizations, and lattice manipulation operations in an effort to establish a possible dataset augmentation strategy for the improvement of classification performance of the ResNet. Using the disordered lattice information originating from the defects, test dataset SADPs were generated by simulating electron diffraction in transmission electron microscopy. The ResNet, pre-trained using SADPs from defect-free materials, showed decreasing but acceptable classification accuracies with increasing degrees of lattice disorder regarding the lattice vibrations and point defects. When tested using the diffraction patterns for strained lattices, the ResNet responded to the changing lattice symmetry when strain levels are relatively high suggesting that it has capability to discern different symmetries induced by large strains. However, the ResNet failed to recognize lattice structure when dislocations and twin boundaries were considered. It is suggested that DCNN architectures be trained over various scenarios including changes in the image feature characteristics in the diffraction patterns related to defects in future developments for improved general classification performances.

Machine learning for classifying narrow-beam electron diffraction data.

As an alternative approach to X-ray crystallography and single-particle cryo-electron microscopy, single-molecule electron diffraction has a better signal-to-noise ratio and the potential to increase the resolution of protein models. This technology requires collection of numerous diffraction patterns, which can lead to congestion of data collection pipelines. However, only a minority of the diffraction data are useful for structure determination because the chances of hitting a protein of interest with a narrow electron beam may be small. This necessitates novel concepts for quick and accurate data selection. For this purpose, a set of machine learning algorithms for diffraction data classification has been implemented and tested. The proposed pre-processing and analysis workflow efficiently distinguished between amorphous ice and carbon support, providing proof of the principle of machine learning based identification of positions of interest. While limited in its current context, this approach exploits inherent characteristics of narrow electron beam diffraction patterns and can be extended for protein data classification and feature extraction.

Identification of crystal symmetry from noisy diffraction patterns by a shape analysis and deep learning

The robust and automated determination of crystal symmetry is of utmost importance in material characterization and analysis. Recent studies have shown that deep learning (DL) methods can effectively reveal the correlations between X-ray or electron-beam diffraction patterns and crystal symmetry. Despite their promise, most of these studies have been limited to identifying relatively few classes into which a target material may be grouped. On the other hand, the DL-based identification of crystal symmetry suffers from a drastic drop in accuracy for problems involving classification into tens or hundreds of symmetry classes (e.g., up to 230 space groups), severely limiting its practical usage. Here, we demonstrate that a combined approach of shaping diffraction patterns and implementing them in a multistream DenseNet (MSDN) substantially improves the accuracy of classification. Even with an imbalanced dataset of 108,658 individual crystals sampled from 72 space groups, our model achieves 80.12 ± 0.09% space group classification accuracy, outperforming conventional benchmark models by 17–27 percentage points (%p). The enhancement can be largely attributed to the pattern shaping strategy, through which the subtle changes in patterns between symmetrically close crystal systems (e.g., monoclinic vs. orthorhombic or trigonal vs. hexagonal) are well differentiated. We additionally find that the MSDN architecture is advantageous for capturing patterns in a richer but less redundant manner relative to conventional convolutional neural networks. The proposed protocols in regard to both input descriptor processing and DL architecture enable accurate space group classification and thus improve the practical usage of the DL approach in crystal symmetry identification.

奈良時代に製作された和同開珎の表面腐食層の微細構造

The microstructures of a corrosion layer formed on the surface of two Japanese oldest coins, Wadoukaichin (surface color: dark brown and green) were studied. Compositions of these coins are that main element is Cu and Sn and traces are Pb, Fe and As. X-ray diffraction patterns show that copper di-hydroxide (Cu(OH)2 2H2O), Cu1.81S and Cu2O are detected on the surface corrosion layer of dark brown sample, and Cu2(CO3)(OH)2 (malachite), Cu3(CO3)2(OH)2 (azurite) and Cu2O are detected on the surface corrosion layer of green sample. Optical microscope images show that the surface corrosion layer of dark brown sample is porous, and blue and green grains on the brown layer are existed on that of green sample. From spectrographs of these grains, the peak wavelength of green grains is matched to that of malachite, but this of blue grains is shifted to long wavelength side of about 40 nm at azurite on an influence of trace contaminations. The corrosion layer has two layers, surface layer is plate-like texture, and under layer is aggregation microstructure from TEM observation. The crystal structure of these two layers is matched to azurite from the electron beam diffraction patterns.

Structural insights into ion conduction of layered double hydroxides with various proportions of trivalent cations

We report herein the structural influence of in-plane cation arrangements on ion conductivities of layered double hydroxides (LDHs). Mg–Al–CO3 and Mg–Ga–CO3 LDHs with different proportions of trivalent cations were synthesized by the co-precipitation method, and we found that their ion conductivities presented a steep increase only at peculiar proportions of trivalent cations. Electron-beam diffraction patterns revealed that LDHs with such proportions had an ordered honeycomb arrangement of di- and trivalent cations in (0001) hydroxide layers. We propose a crucial insight that relates the order/disorder arrangement and ion conductivities for better understanding the nature of LDHs.

A Method for Complete Characterization of the Macroscopic Geometry of Grain Boundaries

In the current paper a new method is presented that determines the five macroscopic parameters of a grain boundary (GB) from electron diffraction patterns and bright field (BF) images in a transmission electron microscope. Usefulness of the method is demonstrated on a set of GBs measured in laser crystallized Si thin films. To characterize the geometry of a grain boundary, we have to determine the misorientation between the neighboring grains, and the direction of the GB-plane. The misorientation is calculated from two convergent beam electron diffraction (CBED) patterns taken on the neighboring grains, and the plane-normal can be calculated from bright field (BF) images of the GB. The thickness of the sample is also needed and it is measured from a CBED pattern taken in two-beam condition. It has been previously observed in metallic thin films, that the GBs can minimize their energy in two alternative ways in thin films: either forming a GB plane with low energy density or minimizing the area of the plane. In the last case the GB plane is almost normal to the surface of the specimen and has a general index. We found, that boundaries with exact Σ3 misorientation generally adopted low energy-density {111} boundary planes and formed coherent twins. The rest of the boundaries adopted general index planes and minimized the surface area of the boundary. The last group included boundaries with small deviation from 3 misorientation, other special high-angle boundaries with-value>3 and boundaries with general misorientation.

Formation of AgCl Cubic Crystals Induced by Shrinkage of Sol–Gel Silica Film

Growth of AgCl cubic crystals at the surface of sol–gel silica films containing AgCl nanocrystals and poly(ethylene glycol)/ethylene glycol as diffusion-controlled media has been successfully demonstrated. Crystals were grown during the shrinking of the film when it was exposed to air at room temperature, and both size and shape of the crystals were found to be dependent on the film thickness. When the film thickness was in the range of 70–250 nm, cubic shaped crystals with size between 120 and 250 nm were produced. These cubes exhibited predominant {200} facets as indicated by the electron beam diffraction patterns. Spherical crystals with size and height between 200 and 400 nm were observed when film thickness was in the range of 250–400 nm, while elliptical crystals were found on films thicker than 400 nm. Crystals were not grown when the film thickness was <25 nm. These results suggest that sol–gel silica films can be utilized as a nanocrystal reservoir to grow crystals of desired size and shape with their locations controlled by the pattern of the film.

Ripples and Layers in Ultrathin MoS2 Membranes

Single-layer molybdenum disulfide (MoS2) is a newly emerging two-dimensional semiconductor with a potentially wide range of applications in the fields of nanoelectronics and energy harvesting. The fact that it can be exfoliated down to single-layer thickness makes MoS2 interesting both for practical applications and for fundamental research, where the structure and crystalline order of ultrathin MoS2 will have a strong influence on electronic, mechanical, and other properties. Here, we report on the transmission electron microscopy study of suspended single- and few-layer MoS2 membranes with thicknesses previously determined using both optical identification and atomic force microscopy. Electron microscopy shows that monolayer MoS2 displays long-range crystalline order, although surface roughening has been observed with ripples which can reach 1 nm in height, just as in the case of graphene, implying that similar mechanisms are responsible for the stability of both two-dimensional materials. The observed ripples could explain the degradation of mobility in MoS2 due to exfoliation. We also find that symmetry breaking due to the reduction of the number of layers results in distinctive features in electron-beam diffraction patterns of single- and multilayer MoS2, which could be used as a method for identifying single layers using only electron microscopy. The isolation of suspended single-layer MoS2 membranes will improve our understanding of two-dimensional systems, their stability, and the interplay between their structures, morphologies, and electrical and mechanical properties.

Quantitative and easy estimation of a crystal bending effect using low-order CBED patterns

The quantitative measurement of a crystal bending effect is performed using low-order zone-axis convergent beam electron diffraction (CBED) patterns. Although the accuracy of the present method is inferior to that of the method of using split higher order Laue zone lines, this method enables us to estimate the crystal bending effect at a region very close to the interface and to easily judge whether the crystal bending effect results in a tensile bend or a compressive bend. As an application of the present method, the crystal bending effect at a region close to the SiGe/Si interface was measured. It was found that the crystal bending effect is due to a thin-foil relaxation of almost 0.3 degrees at a region that is approximately 10 nm away from the interface.

Hydrogenation of Ti–12Mg blending powder during mechanical milling in liquid media containing hydrogen

The authors manufactured the metal hydrides (Ti0·88Mg0·12)H2 and (Ti0·88Mg0·12)H0·7 by a very easy and cheap way using the mechanical milling of Ti–12Mg blending powder in liquid milling media such as isopropyl alcohol, hexane and heptane as hydrogen sources. The powder was analysed by X-ray diffraction (XRD) and convergent beam electron diffraction (CBED) patterns parallel with the chemical analysis by EDX and ion coupled plasma (ICP) measurements. The hydriding reaction began when the mechanically milled particle achieved nanometre size. (Ti0·88Mg0·12)H2 was manufactured in isopropyl alcohol and hexane media, and (Ti0·88Mg0·12)H0·7 in heptane, while the Ti–12Mg blending powder milled in argon atmosphere formed the solid solution of hexagonal close packed (hcp) structure.

Diffraction patterns of artificial two‐dimensional crystals synthesized in situ in an environmental scanning transmission electron microscope

In this study, we demonstrated the use of electron-beam-induced deposition for synthesis of artificial two-dimensional crystals with an in situ scanning transmission electron microscope. The structures were deposited from W(CO)6 in an environmental scanning transmission electron microscope on a 30-nm-thick Si3N4 substrate. We present clear electron beam diffraction patterns taken from those structures. The distance between the diffraction peaks corresponded to the dot spacing in the self-made surface crystal. We propose using these arrays of dots as anchor points for making artificial crystals for diffraction analysis of weakly scattering or beam-sensitive molecules such as proteins.

Crystal orientation distribution in highly oriented diamond films investigated by SEM and TEM

Orientation imaging techniques based on electron backscattering diffraction (EBSD) and convergent beam electron diffraction (CBED) pattern acquisition carried out respectively in the scanning electron microscope (SEM) and in the transmission electron microscope (TEM) were used to characterise the orientation of crystals in two highly oriented diamond (HOD) films with different thicknesses. The substrate surface modification has been also investigated by EBSD. The films were synthesised on Si(100) in a microwave plasma chemical vapor deposition (MPCVD) reactor where nucleation is initiated via an in situ pretreatment consisting in a carburization step followed by a bias enhanced nucleation (BEN). For the thin film EBSD allowed to highlight the presence of alpha-SiC on the silicon substrate as well as to evaluate their epitaxial relationship. The TEM-based system allowed to obtain more accurate misorientation data of the fine grained diamond structure corresponding to the first stages of growth. For the thicker film, EBSD was used to determine the mosaicity at the surface.

Electron microscopy study of the sakhaite from endogenous borate deposit in Hunan, China

Sakhaite specimens from an endogenous borate deposit in Hunan Province, China have been studied in detail by means of transmission electron microscopy (TEM). The selected area electron diffraction (SAED) patterns obtained with different crystal zone axes show that the crystal system is cubic, with a = 1.463 nm. The convergent beam electron diffraction (CBED) patterns indicate that the point group and space group of sakhaite is m3m and Fd3m, respectively. The space group of synthetic sakhaite determined by Chichagov (1974) is F4 1 32. The possible reasons why the space groups of the natural and the synthetic sakhaites are different are discussed.

Transmission Electron Microscopy Study of Y2O3 Nanotips Grown on LaAlO3

Y2O3 is a super refractory oxide with high thermal stability and finds various applications in optics and microelectronic devices. Recently, Eu-activated Y2O3 films attracted much research interest due to its promising applications in flat panel field emission displays. Epitaxial Y2O3:Eu thin films have been grown on LaA1O3(LAO). in this paper we report a transmission electron microscopy (TEM) study of Y2O3 nano tip-structure grown on LAO by pulsed laser deposition using stoichiometric YBa2Cu3O7-σ.a target under a low oxygen pressure. The experimental work was conducted within a JEOL2010F TEM equipped with an ED AX system.Fig.l is a low magnification cross-section TEM image of the nano-tip structure grown on (001) LAO. The lattice parameters of the tips were calculated to be that of Y2O3 by using LAO as a standard for indexing the diffraction patterns. Nano electron beam diffraction patterns from the tips along [100] and [110] axis are shown in Fig.2a and Fig.2b, respectively, in agreement with the simulated ones using bulk Y2O3 structure.

A study of quasicrystal structure factors by using quantitative CBED

The procedure of simulating convergent beam electron diffraction (CBED) pattern of quasicrystals by dynamical theory is described. The simulated patterns are generally coincide with the experimental patterns. The variations of intensity distribution in CBED pattern with the amplitude and phase of the structure factor of quasicrystal are calculated with dynamical theory. The sensitivity of intensity distribution to the structure factor is investigated.

Preparation and properties of boron thin films

Boron thin films were deposited by pyrolysis of decaborane. The substrate was heated up to 1200 °C by infrared irradiation from a halogen lamp. The films were characterized by electron beam diffraction, XPS and electrical measurements. Spots and rings were observed in the electron beam diffraction patterns, which indicates that the films consist of α-rhombohedral boron. The activation energy of the conductivity was 0.1–0.3 eV from R.T. to 500 °C and 1.0–1.4 eV above 500 °C. This demonstrates that infrared radiation is an effective method in obtaining high quality boron films via the pyrolysis of decaborane.

Templating fabrication of platinum nanoparticles and nanowires using the confined mesoporous channels of FSM-16—their structural characterization and catalytic performances in water gas shift reaction

The robust trigonal prismatic Pt 18 and Pt 15 cluster anions are selectively synthesized via [Pt(CO)Cl 3] − in FSM-16 (4.7 nm) and cis-Pt(CO) 2Cl 2 in FSM-16 (2.8 nm) by the reductive carbonylation of H 2PtCl 6 in the confined mesoporous channels of FSM-16, respectively. The Pt cluster anions extracted from the resulting samples in the THF solution by cation metathesis with [PPN]Cl are identified as [Pt 3(CO) 6] 5 2− in FSM-16 (2.8 nm) and [Pt 3(CO) 6] 6 2− in FSM-16 (4.7 nm) by the FTIR and UV–Vis spectroscopic data. The EXAFS and FTIR studies demonstrated that [Pt 3(CO) 6] 5 2−/FSM-16 (2.8 nm) ( ν CO=2078, 1880 cm −1) and [Pt 3(CO) 6] 6 2−/FSM-16 (4.7 nm) ( ν CO=2056, 1879 cm −1) were transformed by heating at 300–343 K into the partially decarbonylated Pt clusters (Pt–Pt; C.N.=7.6, R=2.73 Å) characteristic of a linear CO band ( ν CO=2015 cm −1). According to the EXAFS and TEM observation, Pt carbonyl clusters are eventually converted above 473 K to naked Pt particles of 11 Å (Pt–Pt C.N.=6.7, R=2.72 Å) in FSM-16 (2.8 nm) and of 15 Å (C.N.=7.9, R=2.74 Å) in FSM-16 (4.7 nm). On the other hand, the platinum nanowires [2 and 4 nm (diameter)×50–500 nm (long)] are prepared using the cylindrical mesoporous channels of FSM-16 (2.8 and 4.7 nm) in the templating reduction of H 2PtCl 6 by the exposure to γ-ray or UV-light ( λ max>254 nm) under the atmosphere of 2-propanol and water. It was demonstrated by the TEM observation and EXAFS characterization that the diameters of Pt nanowires formed in FSM-16 are consistent with those of FSM-16 used (2.8 and 4.7 nm), whereas the average lengths of Pt wires are varied by the function of the Pt loading mass and time exposure to γ-ray and UV-light. The observed electron beam diffraction patterns indicate that the Pt wires consist of a single crystal phase of Pt(110) having a lattice fringe of 2.27 Å. It is of interest to find that the Pt nanowires in FSM-16 (2.8 nm and 4.7 nm) exhibited an unique IR spectrum in CO chemisorption and the remarkable activities per surface Pt atom (TOF) for water gas-shift reaction at 323–393 K by 60–90 times larger than those of Pt nanoparticles in FSM-16 and even conventional metal catalysts.

Iterative phase retrieval from kinematic rocking curves in CBED patterns

In the two-beam limit, the intensity distribution in the dark field disc of a convergent beam electron diffraction (CBED) pattern represents a rocking curve mapped across the Bragg condition. For kinematic scattering from crystal planes which are bent or displaced within the illuminated column, the diffracted amplitude and phase is the Fourier transform of a phase function with an exponent proportional to the displacement normal to the planes. Recovery of the phase profile is formally equivalent to the one-dimensional phase retrieval problem for an object function with constant modulus within a compact support, given only the diffracted intensity distribution. In simulations using the error reduction algorithm, the computed solution always converged to the original phase profile. As a practical test, the asymmetric rocking curves diffracted from planes inclined to the surfaces in ion-thinned Si specimens were used as the diffraction constraint, with support width equal to the crystal thickness. The calculated displacement curve was S-shaped, interpreted as a dilation of 1% induced by Ar atoms implanted in the surface layers. The factors which limit the accuracy and spatial resolution for phase recovery along the beam direction are discussed.

Generation of dislocations and subgrains in stir cast AI-6.2Si alloy during semisolid state processing

The generation of a large number of tangled dislocations and subgrain boundaries by dislocation pileups in the stir cast AI-6'2 wt-%Si alloy was observed by a transmission electron microscopy (TEM) study. Misorientations between grains were measured by selected area diffraction (SAD) and convergent beam electron diffraction (CBED) patterns. The density and entanglement of dislocations increased with higher solid fraction (or shear stress) and higher shear rate as a result of increased plastic strain. Longer isothermal stirring time gave rise to higher dislocation density and smaller subgrain size. The meltoff of high angle boundaries, which was suggested by Vogel-Doherty-Cantor's grain boundary melting model, is supported by the observation of subgrain boundaries owing to dislocation pileups during semisolid stirring.

Generation of dislocations and subgrains in semisolid al-6.2wt%Si alloy

Generation of large amount of tangled dislocations and subsequent formation of subgrains by the pile-up of the dislocations in the stircast Al-6.2wt%Si alloy has been observed through a detailed transmission electron microscopy. The misorientations between the subgrains were measured by selected area diffraction patterns and convergent beam electron diffraction (CBED) patterns. The density and the degree of entanglement of dislocations increased with the increase of solid fraction, shear stress and shear rate. Longer isothermal stirring time gave rise to higher dislocation density and smaller subgrain size. The melt-off of high-angle boundaries which was suggested by Vogel-Doherty-Cantor’s grain boundary melting model is supported by the observation of the subgrains within the primary solid particles due to the generation and subsequent pile-up of the extremely large amount of dislocations during semi-solid stirring.