Fluctuation microscopy is an electron microscopy technique sensitive to medium-range order (MRO) in disordered materials. It has been applied to study amorphous germanium and silicon, leading to the conclusion that these materials exhibit more MRO than the conventional continuous random network model for their structure.

As originally proposed by Treacy and Gibson, fluctuation microscopy utilizes mesoscopicresolution (1.5 nm) hollow-cone dark field (HCDF) imaging in a TEM. The normalized variance of such images,

is a measure of the magnitude of fluctuations in the diffracted intensity from mesoscopic volumes of the sample and is sensitive to MRO via the three- and four-body atom distribution functions. Studying V as a function of the diffraction vector magnitude k gives information about the degree of MRO and the internal structure of ordered regions. V as a function of the inverse resolution Q gives information about the characteristic MRO length scale.

We developed a high energy-resolution electron energy-loss spectroscopy (EELS) microscope to investigate electronic structures from specified small specimen areas [1]. EELS gives us information of the dielectric properties and the partial density of states (DOS) of the conduction band. Together with the DOS of the conduction band (unoccupied states), the DOS of the valence band (occupied states) is imperative to understand the whole electronic structure. Unfortunately, the DOS of the valence band cannot be reached directly by EELS.

A partial DOS of the valence band can be obtained by fluorescent X-ray emission spectroscopy (XES). A conventional EELS instrument attached to a TEM usually takes spectra with an energy resolution of about 1 eV, by which the DOS of conduction bands has been studied using ELNES until now. Thus, an energy resolution of about 1 eV is at least necessary for the study of the DOS of the valence band.

Three-dimensional (3D) observation is one of the important trials of future TEM and STEM methods. in the field of biological science, 3D-observation has already become general because the biological materials are usually amorphous and computed tomography (CT) routines can be effectively used. The CT assumes that the image intensity is linearly proportional to the projection of material density. in a crystalline specimen, however, this linearity does not hold due to the dynamical diffraction effect depending on spepimen thickness and incident directions of electron beam. High-angle annular-darkfield STEM (HAADF-STEM) and EDX mapping images could meet this requirement of the linearity because their image intensity are little affected by the dynamical diffraction effect.

Our 3D-STEM is based on a 300 kV FE-TEM/STEM(Hitachi HF-3000) equipped with a specially designed eucentric specimen stage with double spherical fulcra, a newly developed dark/bright field STEM detectors, a 3D-sampling system and a 3D-imaging processor for 3D topography and tomography.

In many new application areas it is necessary to combine atomic resolution imaging with atomic level chemical and crystallographic analysis. The new applications include microbiology, nanotubes, smart materials and sensor initiatives, and the myriad semiconductor fields. Along with these there are many other valuable applications in older and more established technologies such as chemicals, catalysis, pigments and construction materials for chemical plants, airplanes, pipelines, power generation, and other aspects of societal infrastructure. Analysis down to the atomic level will help to solve recalcitrant problems and open up some reluctant opportunities in these fields. They can be very effective in resolving important and long standing issues with huge potential monetary and societal costs and benefits. Key aspects of environmentally sensitive corrosion and pollution control may also require the support of similarly sophisticated imaging and microanalysis.

We think the capabilities of a new generation high performance instrument should include atomic resolution coherent TEM imaging and also incoherent atomic number (Z) contrast high angle annular dark field (HAADF) STEM imaging.

The IBM high resolution STEM project has now reached the point where integration of new subsystems into the instrument is the main activity. During the past three years, we have demonstrated a 50meV EELS resolution using a high brightness electron monochromator incorporated into the electron gun [1, 2, 3], and better than 0.14mn spatial resolution using spherical aberration correction at 120KeV. [4] Recent effort has been to identify and remove many sources of instabilities in the original instrument; to rebuild several existing subsidiary pieces of equipment such as the ADF detector, the EELS spectrometer electron optical coupling, and the EELS CCD detection system; and to set up software and hardware control of each system to allow routine operation.

The new STEM column is summarized in Fig. 1. Briefly, the original VG gun flange has been lowered about 4cm, using a spacer, to allow insertion of the Fringe Field monochromator.

Internal boundaries in crystalline microstructures, like polycrystals, thin-film or multi-layer composites, essentially control many important material properties and physical phenomena, such as mechanical deformation and fracture, transport processes or phase transformations. This influence is originating from the local atomistic and electronic structures at the boundaries. Hence, for a. science-based technological design and optimisation of structural and functional materials it is of paramount importance to characterize and understand the structure, chemistry and bonding at interfaces.

Transmission electron microscopy (TEM) has matured to one of the most successful experimental approaches for the analysis of extended structural defects like dislocations and interfaces. This presentation discusses scientific and technological ingredients of TEM for quantitative characterisations on the microscopic level. For selected model cases of internal interfaces it illustrates recent achievements of both experiment and theory. A variety of interface properties and corresponding TEM techniques for their characterisation are summarized in TABLE 1.

Moiré patterns are commonly formed in multiphase system when diffracting planes of similar spacing and orientation lead to interference effects. They can be used experimentally to evaluate the stress distribution in materials [1], to analyze orientation relationships and latttice strain in diffraction contrast microscopy, or, combined with the related geometrical phase technique, to analyze displacements in high resolution lattice images [2,3]. The interpretation of moiré fringes is often not straightforward due to the elastic interaction between the crystals at the interface and the dynamical nature of electron diffraction [4]. However, if the two lattices are fully relaxed, or if a small precipitate crystal is embedded in a large matrix, moiré patterns can give a simple and direct measure of orientation and lattice constants. in the present work, the moiré technique has been applied to the quantitative analysis of lath-shaped Ge or Ge-Si precipitates in Al with the aim to determine the composition (the Si:Ge ratio) from the lattice parameter indicated by the moiré fringes.

The development of ‘quantitative’ high-resolution transmission electron microscopy (QHRTEM) has led to considerable progress in analyzing and understanding the atomistic structure of grain boundaries and heterointerfaces, which often control relevant macroscopic properties of materials. The following results of recent experimental studies and the comparison of these results with atomistic modeling demonstrate the strength of QHRTEM:

1.Grain Boundaries in SrTiO3. Some technical applications of SrTiO3 ceramics rely on special electric properties of this material, which originate from segregation of electrically active point defects to grain boundaries. QHRTEM can substantially contribute to understanding the correlation between the atomistic structure of grain boundaries and the segregation of point defects to these interfaces. As an example for a grain boundary structure solved by QHRTEM, FIGs. la and b present an experimental HRTEM image of the Σ= 3, (111) grain boundary in SrTiO3 and the corresponding atomistic structure as obtained by QHRTEM.

Oxides, nitrides, carbides and silicides contain more than one type of atoms that gives one extra degree of freedom in the grain boundary and interfacial structure. This degree of freedom is the choice of the boundary plane containing different type of atoms which results of several structural multiplicities in the boundary. Coexistence of the structural multiplicities (or boundary domains) in the interface and grain boundary were reported [1, 2]. The structural information in high frequency domain is usually blurred and lost in the lens system of a high resolution transmission electron microscopes (HRTEM) such that the HRTEM image may not truly represent the real structure. Conventionally, the structures of grain boundary and interface were solved using forward method. The forward method involves matching the simulated images from guessed possible models with the through focal series experimental images. Recently, rereconstruction methods such as focal variation holography [3, 4], tiltbeam holography [5] and Gerchberg-Saxton algorithm [6] provide a backward (direct) way to recover the structural information lost in a modern microscope.

The technique of high-resolution transmission electron microscopy (HREM) produces images that contain information about the atomic structure of the specimen. Within additional, very stringent, constraints, the HREM image can contain information about atomic structure of crystal defects, including grain boundaries and interfaces. to extract this information from the image it is necessary to compare the experimental image with a simulated image calculated based upon an atomic model of the specimen.2 in this comparison, investigators have been aided by the use of quantitative techniques.

Atomistic simulations are often used to predict the atomic structure of crystal defects or to simulate the evolution of dynamic processes in crystals, e.g. radiation effects or dislocation motion and interaction. During the development of new models of interatomic interactions, the predictions of simulations are tested against experimental observations to validate new potentials. Grain boundary structure is a good test because atoms residing in the boundary experience environments (interatomic distances and angles) that are significantly different from those experienced by atoms residing in a perfect crystal lattice site.

The phenomenon of segregation is of long standing scientific interest and has been studied extensively, both theoretically as well as experimentally. For our investigations we have chosen the Σ5 symmetric tilt grain boundaries (STGB) in two face-centered cubic (FCC) metals, aluminum and copper. Both metals were doped with only 1 at% of the impurity species (Cu and Ag). One of our major goals in this study was to investigate the size effect on segregation of an impurity to distinct sites at the grain boundary. Therefore we have selected the Ag as an impurity in Cu and Cu as an impurity in Al. The latter one is of special interest for applications like interconnects in microcircuits, where one of the major controlling factors of electromigration is expected to be the diffusion or segregation of Cu atoms at Al grain boundaries.

The model grain boundaries have been fabricated with ultra-high vacuum diffusion bonding of single crystals which bonds the bicrystals under highly controlled environmental conditions.

α-sialon is the isostructural derivative based on the a-silicon nitride (α-Si3N4) structure by partial replacement of Si by Al and N by O. The structure is considered to be stabilized by the incorporation into the large interstices (r=0.146nm) of a metal like Li, Mg, Ca and most small rareearths. Large rare-earth metal ions have been considered not to be able to enter the interstices in α-sialon structure until several recent successes in accommodating large cerium ions (r=0.103nm). This gives rise to questions of the location and distribution of these large cerium ions within the asialon structure. It has been found that α-sialon grains formed in Ce-Si-Al-O-N system are usually rich in a new structural defect which has been considered to be initiated specifically by the presence of cerium in the lattice. This new intergrown defect is identified as domain boundary in our present study by transmission electron microscopy (TEM).

The understanding of the mechanism of oxidation of Zircaloy materials provides an important support for the corrosion control of the fuel claddings in the light water nuclear reactors. Many investigations are devoted to the study of the oxidation of these materials. One of the important aspects of these studies, is the analysis of the metal-oxide interface, which produces information about the nature of the oxide formed at the interface, at different stages of oxidation and the influence of the oxide structure and morphology on the formation and growth of the oxide.

In the present study, analytical electron microscopy (AEM) and high resolution transmission electron microscopy (HRTEM) are used to examine the metal-oxide interface of an un-irradiated Zircaloy-4 material, oxidized in autoclave, under pressurized water reactor conditions.

The TEM specimen preparation for the interface analysis is an important step of the investigation, since the transverse section required for such observation should be sufficiently thin exactly at the position of the interface.

Epitaxial ZnO thin films were produced on basal plane ScAlMgO4 (SAM) by chemical solution deposition (CSD) (fig. 1). ZnO crystallizes in the wurtzite structure and ScAlMgO4 is isostructural to YbFe2O4, a layered structure shown in fig. 2. The crystallographic orientation relationship was observed by XRD using θ-2θ scans and off-axis φ scans to be (0001)zno ∥ (0001)SAM and zno ∥ SAM. The lattice mismatch between film and substrate is 0,09%.

For TEM observations in cross-sectional geometry two substrates were glued together with the epitaxial film surfaces. Electron transparent specimens were produced by standard methods of mechanical thinning and ion beam etching. The HRTEM investigations were performed on a Philips CM300 FEG electron microscope with Ultra-Twin lens operated at 300 kV. After correction of axial coma and two-fold astigmatism images were recorded on a slow-scan lk × lk CCD camera with a sampling of 0.0193 nm/pixel.

For over 30 years geologists have been trying to better understand antiphase domains (APD) and boundaries (APB) in pigeonite in hopes of using them as markers for the thermal history of the rocks in which they are found. The ability to know the cooling history of igneous rocks is of great interest to geologists and pigeonite has received special attention on this matter because it has exsolution (precipitation) and antiphase domains (APD), both of which can be used as possible thermal markers. APDs in pigeonite arise because of the C2/c → P21/c transformation that occurs upon cooling. When multiple APDs nucleate, grow, and impinge upon one another, they are either in registry or have a translational discrepancy of ½(a+b). The size of the APDs can be used as a qualitative marker of cooling rates, since slowly cooled pigeonites favor large APDs and rapidly cooled pigeonites favor small APDs.

A TiN thin film was grown by the DC-Sputtering technique was characterized by EELS spectroscopy in order to obtain the optical quantities, ε1,ε2 and the refractive index η. Optical properties were obtained analyzing the low loss window located at an energy less to 50 eV respect to the elastic peak. A primary energy of 200 keV and a spot size of 440 nm with an average acquisition time of 5 min were used. The optical properties were calculated using the Kramers - Kronig relations.

Figure 1 shows the PEELS spectra in the low loss region for TiN thin film. As we can see, the most prominent plasmon loss peak characterizing TiN appears at an energy loss of 24 eV. The origin of the loss peak located at 17 eV is interpreted as surface plasmon. in order to obtain a better approximation, we use a near-free electron model and we have compared with the respective optical quantities calculated with the conventional formalism.

Vanadium carbide was purchased as powder at Aldrich with a 99.9% pure. Sample was prepared in the standard method for powder observation in a TEM .EELS experiments were carried out in a Phillips CM 200 STEM equipped with a Gatan 766 PEELS spectrometer. Experimental conditions for acquiring were the follows: a spot size of 500 nm, a chamber length of 400 mm and a detector aperture of 3 mm using a energy dispersion of 0.3 eV/channel with a beam energy of 200 KeV. Acquisition time was around 10 mn. taking an average of 100 spectra.

All the EELS spectra were corrected for background subtraction and multiple scattering deconvolution, before the fine structure analysis. The fine structure range was about 200 eV beyond the C-K edge located at energy loss of 283.8 eV and respect to the VL23-edge located at a energy loss of 513 eV.

Titanium Nitride thin film was obtained by the DC-Sputtering technique and characterized using the transmission EXEELFS formalism . The experiments were performed in a CM 200 transmission electron microscope with a GAT AN 766 PEELS attachment. A primary energy of 200 keV and a spot size of 440 nm with an average acquisition time of 5 min were used. The atomic distribution obtained using EXEELFS was corrected with the respective experimental phase shifts. Atomic parameters were compared with those corrected by EXAFS phase shits and the known crystallographic values.

The energy loss spectra for the sample, around the ionization Ti L23-edge and N K-edge corrected for background and multiple scattering process is shows in Figure 1. N K-edge appears at an energy loss of 400.4 eV and the Ti L23-edge is located at an energy loss of 464.3 eV. These values on energy are shifted 0.6 eV respect to the isolated transitions.

Ceramic of ZrB2 and its composite of the ceramics with SiC are used for ultrahigh temperature protection coatings in re-entry space vehicles because of high melting temperature of ZrB2 (3245°C). Chemical composition and microstructure of the ceramics are critical to thennal shock properties of the ceramics. Especially oxidation of boron and formation of boron oxide will degrade property of thermal resistance. Conventional TEM with Energy dispersive X-ray spectroscopy (EDS) is unable to study light elements like boron. However, parallel-recorded electron energy-loss spectroscopy (EELS) and energy-filtering GIF system associated with field-emission gun (FEG) TEM are able to carry out chemical imaging and chemical information of light elements. Bonding structure and oxidation states of an element (such as B) can be obtained by using EELS.

Figure A is a zero-loss bright-field TEM image showing an interface between ZrB2 and SiC. Figure B is a Plasmon-loss filtered TEM image of the same area. A plasmon peak at about 20 eV was used for this image.

The excellent mechanical properties of Ni based superalloys depend upon the presence of γ’ particles (LI2 structure). Their volume fraction, spatial distribution and size determine the mechanical strength of these alloys. Ni alloys for technological applications make use of large volume fractions of precipitates where processes of coarsening and coalescence take place during service leading in some cases to deterioration of properties. Addition of different alloying elements prevents accelerated coalescence by retarding diffusion and thus improving the mechanical properties of such alloys. Coalescence can also take place under the influence of an applied stress leading to the formation of rafts of the y' phase. For example the microstructure changes during creep deformation, depending on the alloy composition, with the corresponding formation of dislocation networks and rafts of different morphologies [1]. The γ-γ’ interfaces are also different depending on the alloy composition and most likely to the local distribution of alloying elements and their strain fields.