Search

Energy-dispersive X-ray spectroscopy (EDXS) in a transmission electron microscope is frequently used for the chemical analysis of Cu(In,Ga)Se2 (CIGS) solar cells with high spatial resolution. However, the quantification of EDXS data is complicated due to quantification errors and artifacts. This work shows how quantitative EDXS analyses of CIGS-based solar cells with Zn(O,S) buffer and ZnO-based window layers can be significantly improved. For this purpose, CIGS-based solar cells and a reference sample with a stack of Zn(O,S) layers with different [O]/[S] ratios were analyzed. For Zn(O,S), the correction of sample-thickness-dependent absorption of low-energy O–Kα X-rays significantly improves the results of quantitative EDXS. Absorption of characteristic X-rays in CIGS is less relevant. However, for small transmission electron microscopy (TEM) sample thicknesses, artifacts can occur due to material changes by focused-ion-beam (FIB)-based preparation of TEM samples, electron-beam-induced damage, and oxidation of the sample surface. We also show that a Pt-protection layer, deposited on the sample surface before FIB preparation of TEM lamellae, can induce artifacts that can be avoided by first depositing a carbon layer.

Contamination is an undesired side effect in many electron microscopy studies that covers structures of interest and degrades resolution. Although contamination has been studied for decades, open questions remain regarding favorable imaging conditions for contamination minimization and the efficiency of contamination-mitigation strategies. This work focuses on electron-beam-induced carbon contamination in scanning transmission electron microscopy at electron energies of 30 keV and below. A reliable method to measure contamination thicknesses was developed in this work and enables the identification of imaging conditions that minimize contamination. Thin amorphous carbon films were used as test samples. The variation of important imaging parameters shows that the contamination thickness increases with the reduction of the electron energy to about 1 keV but decreases below 1 keV. Contamination increases with the beam current but saturates at high currents. Applying a given dose with a high dose rate reduces contamination. Among the tested contamination-mitigation methods, plasma cleaning and beam showering are most effective. Most experiments in this work were performed with focused scanning illumination. Experiments were also carried out with a stationary defocused beam for comparison with a theoretical contamination model with good agreement between measured and calculated contamination thickness.

Focused-ion-beam/scanning-electron-microscopy (FIB-SEM) tomography has become increasingly important to investigate the three-dimensional (3D) distribution of different phases in inhomogeneous materials. However, quantitative analysis of 3D structures by FIB-SEM tomography requires adequate image segmentation that is typically based on thresholds in the intensity histograms of SEM images. Recently machine learning (ML) segmentation algorithms have emerged with the opportunity to tune the algorithm based on prior knowledge of SEM image contrast. In this work we have performed FIB-SEM tomography of 3D printed nanoporous polymer structures with a special focus on SEM image segmentation and pore size determination. By applying Monte Carlo (MC) simulations, SEM image contrast of the pore/polymer interface was analysed. The performance of several traditional segmentation algorithms on simulated SEM images was found to lead to erroneous results on pore sizes. Training the ML segmentation algorithm by the knowledge obtained from MC simulations yields more reliable segmentation results. Evaluated pore sizes and analysis of further 3D material properties show a strong dependence on the segmentation method, which emphasize the importance of the segmentation process in the 3D reconstruction of FIB-SEM data.

Nanoporous, high-purity magnesium nitride (Mg3N2) was synthesized with a liquid ammonia-based process, for potential applications in optoelectronics, gas separation and catalysis, since these applications require high material purity and crystallinity, which has seldom been demonstrated in the past. One way to evaluate the degree of crystalline near-range order and atomic environment is electron energy-loss spectroscopy (EELS) in a transmission electron microscope. However, there are hardly any data on Mg3N2, which makes identification of electron energy-loss near-edge structure (ELNES) features difficult. Therefore, we have studied nanoporous Mg3N2with EELS in detail in comparison to EELS spectra of bulk Mg3N2, which was analyzed as a reference material. The N-K and Mg-K edges of both materials are similar. Despite having the same crystal structure, however, there are differences in fine-structural features, such as shifts and absences of peaks in the N-K and Mg-K edges of nanoporous Mg3N2. These differences in ELNES are attributed to coordination changes in nanoporous Mg3N2caused by the significantly smaller crystallite size of 2–6 nm compared to the larger (25–125 nm) crystal size in a bulk material.

(Ba0.5Sr0.5)(Co0.8Fe0.2)O3−δ (BSCF) is a promising material with mixed ionic and electronic conductivity which is considered for oxygen separation membranes. Selective improvement of material properties, e.g. oxygen diffusivity or suppression of secondary phase formation, can be achieved by B-site doping. This study is concerned with the formation of Co-oxide precipitates in undoped BSCF at typical homogenization temperatures of 1,000°C, which act as undesirable nucleation sites for other secondary phases in the application-relevant temperature range. Y-doping successfully suppresses Co-oxide formation, whereas only minor improvements are achieved by Sc-doping. To understand the reason for the different behavior of Y and Sc, the lattice sites of dopant cations in BSCF were experimentally determined in this work. Energy-dispersive X-ray spectroscopy in a transmission electron microscope was applied to locate dopant sites exploiting the atom location by channeling enhanced microanalysis technique. It is shown that Sc exclusively occupies B-cation sites, whereas Y is detected on A- and B-cation sites in Y-doped BSCF, although solely B-site doping was intended. A model is presented for the suppression of Co-oxide formation in Y-doped BSCF based on Y double-site occupancy.

A fast method for determination of the Co-valence state by electron energy loss spectroscopy in a transmission electron microscope is presented. We suggest the distance between the Co-L3and Co-L2white-lines as a reliable property for the determination of Co-valence states between 2+ and 3+. The determination of the Co-L2,3white-line distance can be automated and is therefore well suited for the evaluation of large data sets that are collected for line scans and mappings. Data with a low signal-to-noise due to short acquisition times can be processed by applying principal component analysis. The new technique was applied to study the Co-valence state of Ba0.5Sr0.5Co0.8Fe0.2O3-d(BSCF), which is hampered by the superposition of the Ba-M4,5white-lines on the Co-L2,3white-lines. The Co-valence state of the cubic BSCF phase was determined to be 2.2+ (±0.2) after annealing for 100 h at 650°C, compared to an increased valence state of 2.8+ (±0.2) for the hexagonal phase. These results support models that correlate the instability of the cubic BSCF phase with an increased Co-valence state at temperatures below 840°C.

In this study the nanomorphology of P3HT:PC61BM absorber layers of organic solar cells was studied as a function of the processing parameters and for P3HT with different molecular weight. For this purpose we apply scanning transmission electron microscopy (STEM) at low electron energies in a scanning electron microscope. This method exhibits sensitive material contrast in the high-angle annular dark-field (HAADF) mode, which is well suited to distinguish materials with similar densities and mean atomic numbers. The images taken with low-energy HAADF STEM are compared with conventional transmission electron microscopy and atomic force microscopy images to illustrate the capabilities of the different techniques. For the interpretation of the low-energy HAADF STEM images, a semiempirical equation is used to calculate the image intensities. The experiments show that the nanomorphology of the P3HT:PC61BM blends depends strongly on the molecular weight of the P3HT. Low-molecular-weight P3HT forms rod-like domains during annealing. In contrast, only small globular features are visible in samples containing high-molecular-weight P3HT, which do not change significantly after annealing at 150°C up to 30 min.

A promising novel type of electrostatic phase plate for transmission electron microscopy (TEM) is presented. The phase plate consists of a single microcoaxial cable-like rod with its electrode exposed to the undiffracted electrons. The emerging field is used to shift the phase of the undiffracted electrons with respect to diffracted electrons. The design overcomes the drawback of the spatial frequency-blocking ring electrode of the Boersch phase plate. First, experimental phase-contrast images are presented for PbSe and Pt nanoparticles with clearly varying phase contrast, which depends on the applied voltage and resulting phase shift of the unscattered electrons. With the new phase-plate design, we show for the first time the reconstruction of an object wave function based on a series of only three experimental phase-contrast TEM images obtained with an electrostatic phase plate.

Achieving phase contrast for weak-phase objects has been the driving force for the development of physical phase plates (PP) for transmission electron microscopy (TEM). Thin-film PPs have meanwhile reached a well-developed state [1]. However, carbon-based PPs suffer from fast material degradation under intense electron irradiation which has initiated the evaluation of alternative materials [2] and other PP approaches as reviewed in [3]. In contrast to thin-film PPs, electrostatic PPs allow the variation of the phase shift which opens the possibility to optimize phase contrast for different types of specimens and facilitate, e.g., wave-function reconstruction not only of weak-phase objects [4]. Among the different electrostatic PP approaches, the Zach PP [5] appears to be most promising because it minimizes obstructing structures in the back focal plane (BFP) and, hence, artifacts induced by blockade of spatial frequencies. Substantial improvements have been made with respect to contamination and charging which yields phase-contrast TEM images with well-controlled phase shift. This was demonstrated by Frindt et al. [6] where imaging with a Zach PP yielded for the first time in-focus phase-contrast images of weak-phase objects (filamentous actin embedded in vitreous ice) with negligible artifacts.

Extended abstract of a paper presented at Microscopy and Microanalysis 2012 in Phoenix, Arizona, USA, July 29 – August 2, 2012.

Extended abstract of a paper presented at Microscopy and Microanalysis 2012 in Phoenix, Arizona, USA, July 29 – August 2, 2012.

Extended abstract of a paper presented at Microscopy and Microanalysis 2012 in Phoenix, Arizona, USA, July 29 – August 2, 2012.

Extended abstract of a paper presented at Microscopy and Microanalysis 2012 in Phoenix, Arizona, USA, July 29 – August 2, 2012.

Extended abstract of a paper presented at Microscopy and Microanalysis 2010 in Portland, Oregon, USA, August 1 – August 5, 2010.

Extended abstract of a paper presented at Microscopy and Microanalysis 2010 in Portland, Oregon, USA, August 1 – August 5, 2010.

Extended abstract of a paper presented at Microscopy and Microanalysis 2010 in Portland, Oregon, USA, August 1 – August 5, 2010.

Extended abstract of a paper presented at Microscopy and Microanalysis 2010 in Portland, Oregon, USA, August 1 – August 5, 2010.

Extended abstract of a paper presented at Microscopy and Microanalysis 2010 in Portland, Oregon, USA, August 1 – August 5, 2010.

Transmission electron microscopy phase-contrast images taken by amorphous carbon film-based phase plates are affected by the scattering of electrons within the carbon film causing a modification of the image-wave function. Moreover, image artefacts are produced by non-centrosymmetric phase plate designs such as the Hilbert-phase plate. Various methods are presented to correct phase-contrast images with respect to the scattering of electrons and image artefacts induced by phase plates. The proposed techniques are not restricted to weak-phase objects and linear image formation. Phase-contrast images corrected by the presented methods correspond to those taken by an ideal centrosymmetric, matter-free phase plate and are suitable for object-wave reconstruction.

To determine quantitatively the In-concentration, e.g. in InGaAs quantum-well structures, composition evaluation by lattice fringe analysis (CELFA) [1] in transmission electron microscopy (TEM) is a well suited technique. However, applied to quantum dots (QDs) this method cannot account for the three-dimensional shape of QDs buried within a TEM sample. If the QDs are surrounded by a GaAs cap layer and are only contained partially in the TEM sample, the measured In-concentration will be lower than the real one. We suggest here a procedure for the quantification of the In-concentration in InGaAs QDs which accounts for shape and position of a QD within the TEM specimen. The investigated samples were grown by molecular beam epitaxy on GaAs(001) substrates. The GaAs buffer was deposited at a substrate temperature of 500°C, followed by deposition of 2.4 monolayers (ML) InAs at a deposition rate of 0.0057 ML/s. After a growth interruption of 10 s a 28 nm GaAs capping layer was deposited. During the growth process InGaAs QDs with a base length of about 20 nm are formed in the self-organized Stranski-Krastanow growth mode. The relative positions of the QDs within the cross-section sample and the local sample thickness is determined by recording a tilt series of TEM dark-field images. The images are taken by exciting the chemically sensitive (200) beam. The sample is tilted around an axis parallel to the [100] direction in steps of 5° beginning close to the [010]-zone axis. With increasing tilt angle the projection of the thin wetting layer broadens (Figure 1). The relative position of the QD within the specimen can then be derived from its relative position with respect to the projected wetting layer. The local sample thickness is estimated by measuring the width of the projected wetting layer at a given tilt angle and applying simple trigonometry. For the QD shown in Figure 1 we obtain an averaged local sample thickness of 32 nm and with its position within the boundaries of the projected wetting layer it can be concluded that the QD is contained completely within the TEM sample. To determine the QD shape, in addition to the dark-field TEM images, we used plan-view brightfield TEM images which are compared with image simulations [2] and cross-section high angle annular dark-field scanning transmission microscopy (STEM) images. The shape of the QDs was determined to be pyramids with {101} facets - in some cases with truncated tops. For the CELFA analysis cross-section HRTEM lattice fringe images with lattice fringes parallel and perpendicular to the [001]-growth direction were recorded. Artifacts in the CELFA analysis are induced due to the strong bending of the (002) planes. For this reason, only the images taken with the (200) reflection were further analyzed. For the QD in Figure 1 we obtain a maximum Inconcentration of 46%.

Extended abstract of a paper presented at MC 2007, 33rd DGE Conference in Saarbrücken, Germany, September 2 – September 7, 2007

Extended abstract of a paper presented at MC 2007, 33rd DGE Conference in Saarbrücken, Germany, September 2 – September 7, 2007

Extended abstract of a paper presented at MC 2007, 33rd DGE Conference in Saarbrücken, Germany, September 2 – September 7, 2007

Extended abstract of a paper presented at MC 2007, 33rd DGE Conference in Saarbrücken, Germany, September 2 – September 7, 2007

Extended abstract of a paper presented at MC 2007, 33rd DGE Conference in Saarbrücken, Germany, September 2 – September 7, 2007

Extended abstract of a paper presented at MC 2007, 33rd DGE Conference in Saarbrücken, Germany, September 2 – September 7, 2007