The improvement of growth techniques in the characterization of semiconductor nanostructures, has recently resulted in the realization of quasi-zero dimensional semi-conducting devices (quantum dots) of excellent performances and of reproducible quality (1,2). The design and fabrication of these devices strongly depends on the ability to control parameters that influence the quantum confinement namely the shape, dimension and size distribution of the dots. High spatial resolution structural and analytical techniques are crucial to obtain nanoscale information about the shape of the dots, the structural and chemical abruptness of the interfaces, and the composition of the dots (3).

In this paper we show the results of a structural and chemical characterization of In0.5Ga0.5As/GaAs quantum dots grown by low-pressure metal organic chemical vapor deposition (LP-MOCVD) on a (100) GaAs substrate. The growth was performed in a horizontal LP-MOCVD system (AIXTRON 200 AIX).

Quantitative compositional characterization was carried out on a ceramic boundary using atomic resolution high-angle annular dark field scanning transmission electron microscopy (HAADFSTEM), with the aid of HAADF image simulation. HAADF-STEM observation was performed with a JEM-2010F TEM/STEM equipped with a field emission gun and an HAADF detector. And also the image simulation for HAADF was made on a computer by the program that was developed by the authors using Bethe's method.

The material used in this experiment that was provided by TDK Co., is a semiconducting SrTiO3 (STO) ceramic condenser. The bulk substance was formed by the process of sintering in a reduced ambience followed by the reoxidization in the air after surface coating with Bi2O3, PbO and CuO compounds resulting in having a unique structure with Bi diffusion layers at the grain boundary region.

As the demand for quantitative analysis of Transmission Electron Microscope (TEM) images and diffraction patterns increases, questions arise about the reliability of computer simulations using different software packages and algorithms that are used for comparison with the experiment. Here, we compared five multislice programs, which are (with the exception of Cerius2) freely available, with each other and also a Bloch wave code. Since multislice method propagates the electron beams through the crystal slice by slice, the development of differences between the programs can easily be seen by comparing amplitudes and phases of different beams vs. thickness (Pendelloesung plot). We calculated Pendelloesung plots for InP [001] in zone axis and tilted beam orientations and the Si [111] forbidden reflections using the following codes: Cerius2, NCEMSS, Autoslic3, Multis4, EMS5 and an updated Bloch wave code written by J.M. Zuo4.

Cu contamination has become a larger concern as more semiconductor fabrication facilities switch from aluminum to Cu interconnects. The resolution limits of several analytical tools are compared to determine the optimum analysis methods for detecting Cu contamination in semiconductor materials. The elemental detection limits of Secondary Ion Mass Spectrometry (SIMS), Auger Electron Spectrometry (AES), Microcalorimetry and Energy Dispersive Spectrometry (EDS) systems on Scanning Electron Microscopy (SEM), Scanning Transmission Electron Microscopy (STEM), and Transmission Electron Microscopy (TEM) instruments are evaluated for Cu in WSix.

Two different samples were used in this study. One sample has a high uniform concentration (0.9% atomic, 0.7 wt.%) of Cu that was incorporated during the sputter deposition of WSi2. A lower concentration was ion implanted with 63Cu to a dose of lel4 cm-2 and has a peak concentration of lel9 cm"3, or 0.02% atomic.

Interest in rare earth-activated oxide-based phosphor thin films for high-resolution display devices has been growing in the last few decades. However, thin-film phosphors typically have a significantly reduced brightness compared to equivalent powder phosphor materials. Several possible explanations have been suggested for the lower brightness including internal reflection and the small interaction volume between the incident beam and the solid. In this report, we show another factor to be crucial to external radiative efficiency, the porosity of the films. Porosity creates internal surfaces that act as a "dead layer" which decreases the emission efficiency. Using Z-contrast imaging in the scanning transmission electron microscope (STEM) with simultaneous cathodoluminescence (CL) imaging, the dead layer is directly observed, and quantitative accounts for the reduction of luminescent efficiency.

Eu activated Y2O3 thin films with thickness of about 200 nm were deposited on (001) LaAlO3 substrates by laser ablation. TEM analysis of the samples was conducted in a Philips EM-400 electron microscope operated at l00kV.

In the annual dark field (ADF) imaging mode, diffraction intensities inside the ADF detector cutoff angle are discarded. This results a significant loss of the available information. By simultaneous recording the diffraction patterns, we can gain additional information about the specimen, while the ADF image gives the precise position of the probe. This paper describes how to extract the local lattice parameters or to measure the strain field from the zero diffraction disk. The accurate measurement of strain field close to an interface provide the complementary information about the composition and the structure of the interface. An example is the oxygen deficiency in high Tc superconductor of YBa2Cu3O7-δ. The level of oxygen deficiency is too small to be detected by x-ray spectroscopy but affect the lattice parameters significantly that can be experimentally measured [1].

Modern TEM's are not suited for the taking of diffraction-contrast images of defects. Figures in books and papers taken more than a quarter century ago show defects (dislocations, small precipitates and so on) against a uniform background, reflecting constant diffraction conditions across the field of view. Such pictures can not be taken today. Even if the sample is quite flat, TEM images are crossed by “bend contours” which arise, not from the bending of the sample, but from the changing orientation of the beam. In modern TEM's, the objective lens is an immersion lens. The sample sits deep within the magnetic field of the lens and, thus, the electrons are travelling in a helical pattern as they strike the sample. Even with a so-called parallel beam, at different places across the image, the electrons strike the sample at different angles (see figure 1).

The increased demand in the microelectronics industry for higher spatial resolution in the analysis of device defects has focused attention on the use of transmission electron microscopy(TEM). However, in contrast to scanning electron microscopes(SEM), the number of TEM units in the microelectronics industry is still limited. This is because TEM operation and TEM specimen preparation are rather complicated and the results dependent upon operator experience.

A innovative solution using a dedicated ultra thin film evaluation system(HD-2000) based upon a 200kV cold field emission scanning transmission electron microscopy(STEM) technique has been developed[1]. The system has STEM image resolution of 0.24nm and 0.3nm spatial resolution for elemental analysis using energy dispersive X-ray(EDX) spectroscopy with 0.3sr collection solid angle. The system therefore allows materials characterization to be performed at atomic resolution. Another advantage of the system is the ease of use due to newly developed PC-SEM operation software.

In this paper, we use a Bloch wave approach to show that the elastically scattered electrons from neighbouring atomic columns in a zone-axis oriented crystal contribute incoherently to a Z-contrast image regardless of the fact that we have coherent dynamical scattering from a stationary lattice with no absorption. This incoherent nature of the elastic scattering means that through the filtering of the ls-type Bloch states by the detector geometry we approach the Rutherford Z2 dependence of the column intensity for sufficiently large inner detector angle. Thus, annular dark-field (ADF) imaging provides us with a direct incoherent structure image of the atomic-column positions in a zone-axis oriented crystal.

Bloch wave simulations were carried out at 300 kV for the 〈110〉 orientation of InAs (a = 6.06Å) with full, half and empty columns of In atoms. Calculations including 411-beams were carried out for each of the above structural configurations for a single incident partial plane wave vector parallel to the optic axis.

The demand for the higher performance of semiconductor devices, for instance IC functionality, has stimulated industry to further scale the critical dimensions of the semiconductor devices [6]. Also, as parasitic capacitances are reduced in small device structures, the energy loss in these devices is consequently reduced. [1] Achieving the desired performance from such reduced device structures requires optimization of both the interface properties and dopant redistribution processes, such as diffusion and segregation, under various processing conditions. Only through a careful and accurate characterization of these properties on the atomic scale in the electron microscope we can achieve the understanding of the materials interactions during processing necessary to optimize these properties [2].

Conventional imaging and microanalysis techniques in transmission electron microscopy (TEM), such as phase contrast imaging and energy dispersive X-ray spectroscopy (EDS), lack either the spatial resolution or require extensive simulations and through focal series reconstructions to reveal the structure and composition of such interfaces/doping layers (and in the case of simulations may still not give a unique solution).

Recent success of spherical aberration (Cs) correctors for TEM and STEM opened a new generation of microscopy. In order to achieve sub Å resolution in experiment, however, it is necessary to use a monochromator to reduce the effect of chromatic aberration due to the energy spread of the electron source. The monochromator is also necessary for the high energy-resolution EELS (HREELS). Three microscopes - sub Å TEM, 1 Å STEM and HREELS - and those combinations are considered to bring the high performance electron microscopes of the next generation. Among those three, 1 Å STEM has the highest technical difficulties from the point of view of electron optics instrumentation. In this investigation, we will describe the main electron optical problems, which we should solve within a few years, to achieve 1 Å STEM.

The main reason, why 1 Å STEM is so difficult, is the spread of the beam due to the second order aberrations of the energy filter used as the monochromator.

Using parallel beam illumination, electron beam-induced structural transformation has been studied as radiation damages for decades. Recently, however, more attention may need to be paid to beam-induced chemical transformation which happens under illumination by a finely focused beam. Such a small probe is often used in analytical electron microscopy, particularly the core-electron energy loss spectroscopy (EELS). Since the signal for the spectroscopy is very weak long radiation is required and makes the induced chemical reactions happen. Therefore, without attentions to the reactions, the acquired data could lead to misinterpretation of specimen properties. In the present study, using a single crystal calcium fluoride (CaF2), we demonstrated by time-resolved EELS that radiation with a nanometer electron probe could induce chemical reactions between the specimen and the typical residual gases in a microscope column.

The thin films of CaF2 were fabricated by vaporizing flakes of single crystal CaF2 with an electron beam and by dissolving the vapors onto a (001) surface of sodium chloride (NaCl) heated at 300°C.

Recently there has been considerable interest in developing nanometer- and sub-nanometer-scale electronic devices. Of particular interest in this regard is whether individual molecules or molecular complexes can be employed as electronic device elements. Aviram et al. have reported switching and rectification in an organic thin film. More recently, Potember et al. have shown a field-induced conductance transition on a 500 nm scale, but did not demonstrate local reversibility of the transition. The reverse transition was induced only by application of a broad laser pulse or heat. We have observed and replicated reversible conductance transitions in a fully-organic crystalline complex, on a scale close to the dimensions of the unit cell.

A crystalline thin-film organic complex of 3-nitrobenzal malononitrile and 1,4-phenylenediamine (NBMN-pDA), exhibits reversible conductance transitions on the sub-nanometer scale when exposed to local electric field pulses from an STM tip.

FeAl alloys with 40 to 50 at % Al have an impressive combination of oxidation and sulfidation resistance, low density, and low cost and are thus considered as possible substitutes for stainless steels. However, before these materials can be used in structural applications, the fracture characteristics of the alloy's grain boundaries need to be improved. Small additions of carbon and boron have been considered and the resulting properties have been reported. Boron is found to prevent intergranular fracture in iron-rich alloys and to mitigate environmental embrittlement although the improvement is not as dramatic as is seen with Ni3Al alloyed with boron. Boron segregates to the grain boundaries in Fe-Al alloys and boride precipitates at the boundaries and within the grains have been observed.

In this paper, we report the composition and atomic arrangement in a recently discovered phase (designated v phase) found in Fe-Al alloys containing small boron additions using a combination of Z-contrast imaging and electron energy-loss spectroscopy.

The energy-loss spectrum of transmitted electrons contains a wealth of information regarding the physical, chemical and electronic properties of the medium under analysis. It provides a powerful means for materials characterisation in the TEM by use of electron energy-loss spectroscopy (EELS) or its spatially parallel counterpart, energy-selective imaging (ESI). Essentially, both analyses probe the same core-loss information, recording transmitted intensity / as a function of energy-loss E and spatial position x, y, to yield a three-dimensional data set I(E, x, y). Acquisition of an extended series of energy-selected images across the energy-loss range of interest has been shown to provide useful spectral as well as spatial information, with the resolution of extracted ‘image-spectra’ being determined by the energy interval between acquisitions and the width of the energy-selecting slit, as illustrated in Figure la . This mode of analysis, termed ‘image-spectroscopy’ is directly analogous to spectrum-imaging in the STEM, and offers many advantages over conventional two- or three-window elemental mapping techniques .

Crystal structure refinement by the convergent-beam electron diffraction (CBED) method has the following advantages: (i) Nanometer-size crystal structure refinement: CBED patterns can be obtained from specimen areas of a few nanometer in diameter, (ii) Dynamical diffraction effect: the CBED intensities contain phase information of crystal structure factors through the strong dynamical effect. (iii) Site-selective analysis: the use of Bloch states formed by incident electrons allows structure determination weighted for specific atom sites. Energy-filtering (EF) to revome inelastically scattered background is now getting necessary for such quantitative analysis of CBED patterns. Especially for the determination of atom positions and Debye-Waller factors, the use of energy-filtered higher-order Laue zone (HOLZ) CBED patterns is essential because small displacements of atoms can be sensitively detected using HOLZ reflections with large reciprocal vectors. For this purpose, we developed a new fl-filter transmission electron microscope (JEM-2010FEF) with a high acceptance angle [2] and a new analysis program “mbfit” to refine structural parameters using many-beam Bloch-wave calculations and least-square fitting [1].

Many of the electrical, mechanical, and electrochemical properties of diamond-like carbon (DLC) coatings are determined by the amount of sp3 bonding. Correlations of bonding and properties are important in efforts to optimize coatings for specific applications. DLC thin films were grown at LANL by plasma deposition and subsequently ion implanted with N or Ar. Cross-sectioned TEM specimens were characterized with a Gatan imaging filter (GIF) interfaced to a LaB6 Philips CM30T at the ORNL SHaRE User Facility. For elemental distribution maps the following conditions were used: incident beam divergence α = 2.9 mrad, collection angle β = 4.8 mrad, slit width ΔE = 30 eV, exposure times of typically 5 s, 2x-binned 512×512 images, and TEM magnification ∼2000. The standard 3-window method was used for producing elemental maps with AE-r background extrapolation; 2-window jump-ratio images were also produced.

When acquiring energy-filtered TEM (EFTEM) images of a crystalline material, the detrimental effects of diffraction contrast can often be seen in raw energy-filtered images (EFI) (i.e., pre-edge and post-edge images), jump-ratio images and elemental maps as residual diffraction contrast. Residual diffraction contrast occurs in raw EFI because of plural scattering (i.e., inelastic-elastic and elastic-inelastic electron scattering) and in jump-ratio images and elemental maps because background removal procedures often are unable to completely account for intensity changes due to dynamical effects (elastic scattering) that occur between pre-edge and post-edge images acquired at different energy losses.

It is demonstrated in these experiments that, when examining a planar interface, EFTEM images have increased compositional contrast and decreased residual diffraction contrast when the sample is oriented so that the interface is parallel to the electron beam, but not directly on a zone axis.

Spectroscopy and imaging techniques based on electron energy-losses (EELS), which are accessible through energy-filtering transmission electron microscopes (EFTEMs), have proven to be important tools in both materials and life science investigations.

The two most widely used techniques on commercially available EFTEMs are elastic imaging and elemental mapping. Elastic imaging enhances image resolution and contrast by extracting the zero-loss signal and eliminating the inelastic background, whereas elemental mapping, which involves signals coming from element-specific inner-shell ionization edges, is employed to form two dimensional elemental distribution images. In both cases relatively large energy windows of a range of 10 to 30eVare typically used to form energy-filtered images with usually low to moderately high magnifications.

There is however much more information available in an EELS spectrum, which is contained in the detailed fine structure within 0-20eV of a core excitation edge (ELNES) or in the very low energy-loss up to 5eV.

Electron energy loss spectrum-imaging (EELSI) in the energy filtering transmission electron microscope (EFTEM) can provide more accurate analysis of elemental distributions than that obtainable by the standard two-window or three-window background subtraction techniques. Spectra containing many channels can be extracted from regions of interest and analyzed using established methods for quantitation. For example, the pre-edge background can be fitted by an inverse power law and subtracted from the post-edge spectrum. EELSI in the EFTEM is often superior to spectrum-imaging in the scanning transmission electron microscope for mapping specimen regions of size greater than 1 μm. This is due the much larger total beam current that is available at the specimen in a fixed-beam microscope relative to a scanned-beam microscope. Our aim here is demonstrate the advantages of such EELSI measurements for analysis of biological specimens. However, we also indicate some potential pitfalls in acquiring elemental maps in the EFTEM, which can be attributed to specimen instabilities during the acquisition.