Comparison of intensity distributions in tomograms from BF TEM,ADF STEM,HAADF STEM,and calculated tilt series

The three-dimensional (3D) morphology of a nanometer-sized object can be obtained using electron tomography. Variations in composition or density of the object cause variations in the reconstructed intensity. When imaging homogeneous objects, variations in reconstructed intensity are caused by the imaging technique, imaging conditions, and reconstruction. In this paper, we describe data acquisition, image processing, and 3D reconstruction to obtain and compare tomograms of magnetite crystals from bright field (BF) transmission electron microscopy (TEM), annular dark-field (ADF) scanning transmission electron microscopy (STEM), and high-angle annular dark field (HAADF) STEM tilt series. We use histograms, which plot the number of volume elements (voxels) at a given intensity vs. the intensity, to measure and quantitatively compare intensity distributions among different tomograms. In combination with numerical simulations, we determine the influence of maximum tilt angle, tilt increment, contrast changes with tilt (diffraction contrast), and the signal-to-noise ratio (SNR) as well as the choice of the reconstruction approach (weighted backprojection (WB) and sequential iterative reconstruction technique (SIRT)) on the histogram. We conclude that because ADF and HAADF STEM techniques are less affected by diffraction, and because they have a higher SNR than BF TEM, they are better suited for tomography of nanometer-sized crystals.     

Monte Carlo electron-trajectory simulations in bright-field and dark-field STEM: Implications for tomography of thick biological sections

A Monte Carlo electron-trajectory calculation has been implemented to assess the optimal detector configuration for scanning transmission electron microscopy (STEM) tomography of thick biological sections. By modeling specimens containing 2 and 3 at% osmium in a carbon matrix, it was found that for 1-μm-thick samples the bright-field (BF) and annular dark-field (ADF) signals give similar contrast and signal-to-noise ratio provided the ADF inner angle and BF outer angle are chosen optimally. Spatial resolution in STEM imaging of thick sections is compromised by multiple elastic scattering which results in a spread of scattering angles and thus a spread in lateral distances of the electrons leaving the bottom surface. However, the simulations reveal that a large fraction of these multiply scattered electrons are excluded from the BF detector, which results in higher spatial resolution in BF than in high-angle ADF images for objects situated towards the bottom of the sample. The calculations imply that STEM electron tomography of thick sections should be performed using a BF rather than an ADF detector. This advantage was verified by recording simultaneous BF and high-angle ADF STEM tomographic tilt series from a stained 600-nm-thick section of C. elegans. It was found that loss of spatial resolution occurred markedly at the bottom surface of the specimen in the ADF STEM but significantly less in the BF STEM tomographic reconstruction. Our results indicate that it might be feasible to use BF STEM tomography to determine the 3D structure of whole eukaryotic microorganisms prepared by freeze-substitution, embedding, and sectioning.     

Symmetries in BF and HAADF STEM image calculations

Reductions in bright-field (BF) scanning transmission electron microscopy (STEM) and high-angle annular dark-field (HAADF) STEM image calculations with the aid of Bloch wave symmetry are discussed under assumptions that an absorption potential is written by a local potential and a zero-order Laue zone lies parallel to the crystal surface. Translational symmetry allows us to take only partial incident beams in the first Brillouin zone instead of enormous number of partial incident beams in a large convergent disk. Two dimensional point group confines partial incident beams to an irreducible area in addition to factoring a dispersion matrix into noninteracting submatrices on a high symmetry line using the projection operator. The drastic reductions in computing time and memory enable us to readily calculate various BF STEM and HAADF STEM images. The validity and accuracy are demonstrated in comparisons with high resolution experimental BF STEM and HAADF STEM images.     

STEM tomography for thick biological specimens

Scanning transmission electron microscopy (STEM) tomography was applied to biological specimens such as yeast cells, HEK293 cells and primary culture neurons. These cells, which were embedded in a resin, were cut into 1-microm-thick sections. STEM tomography offers several important advantages including: (1) it is effective even for thick specimens, (2) 'dynamic focusing', (3) ease of using an annular dark field (ADF) mode and (4) linear contrasts. It has become evident that STEM tomography offers significant advantages for the observation of thick specimens. By employing STEM tomography, even a 1-microm-thick specimen (which is difficult to observe by conventional transmission electron microscopy (TEM)) was successfully analyzed in three dimensions. The specimen was tilted up to 73 degrees during data acquisition. At a large tilt angle, the specimen thicknesses increase dramatically. In order to observe such thick specimens, we introduced a special small condenser aperture that reduces the collection angle of the STEM probe. The specimen damage caused by the convergent electron beam was expected to be the most serious problem; however, the damage in STEM was actually smaller than that in TEM. In this study, the irradiation damage caused by TEM- and STEM-tomography in biological specimens was quantitatively compared.     

Characterization of precipitates size distribution: validation of low‐voltage STEM

The size distribution of second phase precipitates is frequently determined using conventional transmission electron microscopy (CTEM). However, other techniques, which present different advantages, can also be used for this purpose. In this paper, we focus on high angle annular dark field (HAADF) in TEM and scanning TEM (STEM) in scanning electron microscopy (SEM) imaging modes. The mentioned techniques will be first described, then compared to more conventional ones for the measurement of carbides size distribution in two FeCV and FeCVNb model alloys. This comparative study shows that STEM in SEM, a technique much easier to undertake compared to TEM, is perfectly adapted for size distribution measurements of second phase particles, with sizes ranging between 5 and 200 nm in these systems.     

Effects of specimen tilt in ADF-STEM imaging of a-Si/c-Si interfaces

Annular dark field scanning transmission electron microscopy (ADF-STEM) imaging of a crystal depends strongly on specimen orientation, but for an amorphous sample it is insensitive to orientation changes. To fully investigate the effects of specimen tilt, an interface of amorphous Si (a-Si) and crystalline Si (c-Si) was rotated systematically off a zone axis in a STEM equipped with low-angle ADF (LAADF) and high-angle ADF (HAADF) detectors. The change of relative intensity across the interface shows very different trends in the LAADF and the HAADF images upon tilting. More importantly, it is found that the HAADF signal decreases much more rapidly when tilted off a zone axis than does the LAADF signal. The high-resolution lattice fringes also disappear much faster in the HAADF image than in the LAADF image. These trends reflect the fact that the channeling peaks that are responsible for scattering into the HAADF detector decrease more quickly upon tilting than the lower angle scattering to the LAADF detector does.     

High-angle annular dark-field STEM imaging of immunogold labels

High-angle annular dark-field (HAADF) STEM imaging is a sensitive and efficient technique for detecting immunogold labels. Larger (5–15 nm) gold labels can be distinguished clearly from the heavy metal stain on tissue sections, but for smaller (<5 nm) labels the distinction is less clear. It is not possible to differentiate between ultra-small (<1–3 nm) labels and stain, but the HAADF image shows sufficient contrast so specimens can remain unstained. On a TEM/STEM equipped with a LaB₆ filament the smallest labels that are detectable are 1–2 nm. Efficient detection of ultra-small labels requires a field emission microscope.     

Characterization of the μ and P phase precipitates in the CMSX‐4 single crystal superalloy

A combination of scanning electron microscopy (SEM), transmission electron microscopy (TEM) and scanning‐transmission electron microscopy (STEM) using high‐angle annular‐dark‐field (HAADF) imaging, focussed ion beam‐ scanning electron microscopy (FIB‐SEM) tomography, selected area electron diffraction with beam precession (PED), as well as spatially resolved energy‐dispersive X‐ray spectroscopy (EDS) and electron energy loss spectroscopy (EELS), was used to investigate topologically close‐packed (TCP) phases, occurring in the CMSX‐4 superalloy subjected to high temperature annealing and creep deformation. Structural and chemical analyses were performed to identify the TCP phases and provide information concerning the compositional partitioning of elements between them. The results of SEM and FIB‐SEM tomography revealed the presence of merged TCP particles, which were identified by TEM and PED analysis as coprecipitates of the μ and P phases. Inside the TCP particles that were several micrometres in size, platelets of alternating μ and P phases of nanometric width were found. The combination of STEM‐HAADF imaging with spatially resolved EDS and EELS microanalysis allowed determination of the significant partitioning of the constituent elements between the μ and P phases.     

Accurate structure determination from image reconstruction in ADF STEM

Annular dark-field (ADF) imaging in a scanning transmission electron microscope results in direct structure images of the atomic configuration of the specimen. Since such images are almost perfectly incoherent they can be treated as a convolution between a point-spread function, which is simply the intensity of the illuminating electron probe, and a sharply peaked object function that represents the projected structure of the specimen. Knowledge of the object function for an image region of perfect crystal allows the point-spread function to be directly determined for that image. We examine how the object function for an image can then be reconstructed using a Wiener filter, the CLEAN algorithm and a maximum entropy reconstruction. Prior information is required to perform a reconstruction, and we discuss what nature of prior information is suitable for ADF imaging.     

A versatile double aberration-corrected, energy filtered HREM/STEM for materials science

A HREM/STEM incorporating aberration correctors in both the probe-forming and imaging lenses has been installed at Oxford University. This unique instrument is also equipped with an in-column energy-loss (Omega-type) filter, HAADF detectors above and beneath the filter, and an EDX system. Initial tests have shown it to be capable of approximately 0.1 nm resolution in both TEM and HAADF STEM imaging modes. Some examples of applications are finally presented.     

An expanded approach to noise reduction from high-resolution STEM images based on the maximum entropy method

An expanded use of the maximum entropy method (MEM) is suggested to reduce noise from an experimental high-angle annular dark-field (HAADF) scanning transmission electron microscope (STEM) image. The MEM is combined with an estimate of the standard deviation of noise from an experimental HAADF STEM image and low-pass filtering using the information limit for an incoherent STEM image. Consequently, the present method has just one parameter of a Lagrange multiplier. It is demonstrated that the present method can reduce noise efficiently in high-resolution HAADF STEM images.     

Morphology visualization of irregular shape bacteria by electron holography and tomography

In the current work, irregular morphology of Staphylococcus aureus bacteria has been visualized by phase retrieval employing off‐axis electron holography (EH) and 3D reconstruction electron tomography using high‐angle annular dark field scanning transmission electron microscopy (HAADF‐STEM). Bacteria interacting with gold nanoparticles (AuNP) acquired a shrunken or irregular shape due to air dehydration processing. STEM imaging shows the attachment of AuNP on the surface of cells and suggests an irregular 3D morphology of the specimen. The phase reconstruction demonstrates that off‐axis electron holography can reveal with a single hologram the morphology of the specimen and the distribution of the functionalized AuNPs. In addition, EH reduces significantly the acquisition time and the cumulative radiation damage (in three orders of magnitude) over biological samples in comparison with multiple tilted electron expositions intrinsic to electron tomography, as well as the processing time and the reconstruction artifacts that may arise during tomogram reconstruction.     

Deconvolution processing of HAADF STEM images

A deconvolution processing of high-resolution high-angle annular dark field (HAADF) scanning transmission electron microscopy (STEM) images, combined with maximum entropy method, is applied to two experimental [0 11]-Si images; one having unresolved dumbbells and the other having resolved dumbbells and artificial bright spots. The deconvoluted images for these images show bright spots corresponding to the projected atomic columns and no artificial bright spots. Thus, the deconvolution processing provides almost a real projected atomic structure by eliminating effects of the probe function from HAADF STEM images.     

IMAGE-WARP: a real-space restoration method for high-resolution STEM images using quantitative HRTEM analysis

We have developed a new method for processing distorted high-resolution scanning transmission electron microscopy (STEM) images. The method is based on finding the displaced vertices in the experimental STEM image and warping to geometrically correct reference grid of the object. As a reference grid for warping a structural model obtained using a high-resolution transmission electron microscopy (HRTEM) analysis of the area of interest is utilised. Combined with quantitative HRTEM analysis the IMAGE-WARP method provides a real-space restoration of high-resolution high-angle annular dark-field (HAADF) STEM images without affecting the original Z-contrast information. The method can be applied to extract valuable compositional atomic-column data from any HAADF-STEM image of any kind of bulk crystals with local occupancy or chemistry fluctuations, stacking faults, special grain boundaries or interfaces, for which we have an available structural model. After the warping, distortion-corrected images can be further enhanced using conventional image-filtering techniques, and finally quantified with HAADF-STEM image simulations. The applicability of the IMAGE-WARP method was illustrated using experimental HAADF-STEM images of a strontium titanate crystal disrupted with a Ruddlesden-Popper-type antiphase boundary.     

High-Angle annular dark-field imaging on a tem/stem system

The technique of high-angle annular dark-field (HAADF) imaging, which is highly sensitive to atomic-number contrast, can be performed on TEM/STEM systems using the standard annular dark-field detector. For optimum HAADF imaging, the TEM/STEM must have a high maximum diffraction angle, small minimum camera length, and a descanning facility. The sensitivity of the technique is demonstrated to be about 10⁵ to 10⁶ times higher than energy-dispersive X-ray spectroscopy. Examples are shown from semiconductor, catalysis, ceramics, and particle analysis applications.     

Bloch-wave-based STEM image simulation with layer-by-layer representation

In a dynamical STEM image simulation by the Bloch-wave method, Allen et al. formulated a framework for calculating the cross-section for any incoherent scattering process from the inelastic scattering coefficients: thermal diffuse scattering (TDS) for high-angle annular dark-field (HAADF) and back-scattered electron (BSE) STEM, and ionization for electron energy-loss spectroscopy (EELS) and energy-dispersive X-ray spectroscopy (EDX) STEM. Furthermore, their method employed a skilful approach for deriving the excitation amplitude and block diagonalization in the eigenvalue equation. In the present work, we extend their scheme to a layer-by-layer representation for application to inhomogeneous crystals that include precipitates, defects and atomic displacement. Calculations for a multi-layer sample of Si–Sb–Si were performed by multiplying Allen et al.'s block-diagonalized matrices. Electron intensities within the sample and EDX STEM images, as an example of the inelastic scattering, were calculated at various conditions. From the calculations, 3-dimensional STEM analysis was considered.     

Imaging of high-angle annular dark-field scanning transmission electron microscopy and observations of GaN-based violet laser diodes

The first part of this paper is devoted to physics, to explain high‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) imaging and to interpret why HAADF‐STEM imaging is incoherent, instructing a strict definition of interference and coherence of electron waves. Next, we present our recent investigations of InGaN/GaN multiple quantum wells and AlGaN/GaN strained‐layer superlattice claddings in GaN‐based violet laser diodes, which have been performed by HAADF‐STEM and high‐resolution field‐emission gun scanning electron microscopy.     

Optimization of STEM tomography acquisition — A comparison of convergent beam and parallel beam STEM tomography

In this paper two imaging modes in a state-of-the-art scanning transmission electron microscope (STEM) are compared: conventional STEM with a convergent beam (referred to as nanoprobe) and STEM with a parallel beam (referred to as microprobe). The effect and influence of both modes with respect to their depth of field are investigated. Tomograms of a human white blood cell (hemophagocytes) are acquired, aligned, and evaluated. It is shown that STEM using a parallel beam produces tomograms with fewer distortions and artifacts that allows resolving finer features. Microprobe STEM tomography is advantageous especially in life science, when semi-thin sections (approximately 0.5 μm thick) of biological samples are imaged at relatively low magnification with a large field of view.     

A practical approach for STEM image simulation based on the FFT multislice method

It has been demonstrated that a high-angle annular dark-field (HAADF) STEM technique gives an image resolving atomic columns. Due to the diffusion of this technique and an improvement of its resolution, a practical procedure for image simulation becomes important for a quantitative interpretation of the HAADF image. In this report a new practical scheme for a STEM image simulation is developed based on the FFT multislice algorithm. Here, a HAADF intensity due to thermal diffuse scattering (TDS) is calculated from the absorptive potential corresponding to high-angle TDS and the wave function equivalent to the propagating probe within the sample. Contrary to the commonly used Bloch wave method, a coherent bright-field intensity and a coherent HAADF intensity are also obtained straightforwardly. The HAADF image contrast calculated for GaAs is not simply proportional to Z2 as expected from the Rutherford scattering at high-angle, and the As/Ga contrast ratio depends on the specimen thickness. This suggests that the generation of the HAADF signal is appreciably affected by the coherent dynamical scattering. The developed procedure here will have a definitive advantage over the Bloch wave approach for simulating the HAADF images expected from a defect and interface or amorphous materials, and also the HAADF image obtained by using a Cs-corrected microscope. This is because the former requires a huge super cell, while the latter needs a large objective aperture including a large number of incident beam directions.