Disordered or ‘glassy’ materials are of significant technological value. Their structure can be investigated by diffraction (neutron, X-ray, electron) techniques. When the available volume is very small (eg thin films, non-homogeneous structures) only electrons are suitable. Cockayne and McKenzie developed a technique of RDF analysis using energy filtered electron diffraction which gives information on structures of amorphous materials through the reduced intensity function ϕ(s) and the reduced density function G(r). This technique has been applied successfully to many systems.

A major limitation to the use of G(r) for characterisation of amorphous materials arises from the fact that, for electron diffraction, the G(r) measured is a combination of all partial radial distribution functions for each species in the solid. In addition, in practice there are many different models which will equally well fit the data. In order to overcome these difficulties, refinement needs to be done using a realistic starting atomic configuration. For this reason, we have produced starting models using ab-initio molecular dynamics (Car-Parrinello scheme).

Major advances in the field of in situ environmental high resolution electron microscopy (EHREM) probe selective oxidation catalysis directly on the atomic scale. Dynamic gas-solid surface interactions are studied in real-time and under realistic reaction conditions to unravel atomic level insights into active sites and structure-property relationships in vital chemical and technological processes [1-3]. The recent developments include a pioneering approach with the controlled environmental cell (ECELL) facilities permanently mounted inside the EHREM [2]. Accessories have been added for simultaneous structural and compositional analyses of the reactor contents in real-time, and using atomic resolution imaging with transmission electron diffraction and parallel electron energy loss spectroscopy (PEELS). We are now developing innovative experimental methods that include very high temperature studies and combining in the same instrument facilities for both EHREM and environmental-SEM (ESEM), (EHREM-ESEM), with attractive possibilities for studying the science of selective oxidation catalysis.

Alkane Catalysis, Chlorofluorocarbons and Nanotubes:

In the domain of transition metal based oxides, discoveries of fundamental mechanisms underlying selective catalyzation have come from EHREM studies.

The pivotal role played by Archie Howie in the development of many areas of electron microscopy is universary acknowledged.Here I would like to highlight his contribution to the quantitative description of secondary excitations, which was an important influence on the development of Z-contrast imaging in zone-axis crystals. Secondary excitations are those such as x-ray emission which occur following a primary scattering event, in this case excitation of inner shell electrons. The first important concept to be realized by Archie was that dynamical diffraction and channeling are different manifestations of the same physical effect, namely, the multiple scattering of electrons within a crystal. Second was the realization that processes which are localized within the unit cell will show a dependence on diffraction conditions, such as incident beam orientation, and could therefore be described quantitatively using dynamical diffraction theory. Precisely the same theory was used to describe the orientation dependence of cathodoluminescence

The development of the STEM for high resolution imaging was of course due primarily to Crewe and coworkers, with an annular detector to allow efficient detection of elastic scattering over a wide angular range.

Carbon tubes or spheres synthesized by arc-discharge are usually mixed with other byproducts, prohibiting direct measurements of their physical properties by the well established optical techniques because a large quantity of pure specimen is required. Electron energy-loss spectroscopy (EELS) is a unique technique that can be applied to probe the electronic structure of a single carbon tube or sphere. In this paper, the classical dielectric response theory is applied to calculate the EELS spectra acquired from a graphitic carbon sphere at various impact parameters. Graphite is an anisotropic dielectric medium whose dielectric function is described by a tensor. A graphitic carbon sphere is composed of concentric graphitic shells whose dielectric tensor in the spherical geometry, under the local response approximation, is given by (Figure 1)

where are the dielectric function of graphite for an electric field perpendicular and parallel, respectively, to the c axis. In the non-relativistic approximation, the surface excitation is calculated by [1]

EELS in the low loss region of the spectra (< 50eV) provides information on excitations of outer shell electrons and thus the electronic structure of a specimen material which determines its optical properties. In this work dedicated EELS methods for the experimental acquisition and analysis of spectra are described which give improved information about the electronic structure near the bandgap region at a spatial resolution in the range of nanometers. For this purpose we made use of a cold field emission STEM equipped with a dedicated EELS system. This device provides a subnanometer electron probe and offers an energy resolution of 0.35 eV. Application of suitable deconvolution routines for removal of the zero loss peak extracts information on the closest bandgap region while Kramers-Kronig transformation deduces the dielectric properties from the measured energy loss function. These methods have been applied to characterize the optical properties of wide-bandgap materials for the case of group Ill-nitride compounds which are currently the most promising material for applications on optoelectronic devices working in the blue and ultraviolet spectral range.

Crystalline interfaces associated with lattice translations such as stacking-faults (SFs) and twin boundaries may be characterized in the transmission electron microscope (TEM) by oscillatory fringe contrast known as SF, or rigid-body-translation, fringes. Such fringe contrast is usually obtained using weak-beam bright-field (BF) or dark-field (DF) conditions with ideally only one reflection, g, excited. The nature of the SF contrast has been explained by dynamical theory of electron diffraction in terms of the Howie-Whelan equations [1]. The fringe contrast arises from an interference caused by a phase shift the diffracted electron beams, undergo upon propagating across the interface. The phase shift, α, is proportional to the projection of the translation vector, R, onto the diffraction vector, g (i.e. α # g.R). The direction of the translation state can be determined explicitly from the SF contrast in a manner analogous to the analysis of dislocation Burgers vector.

An analysis of SF fringes with respect to diffraction theory permits the magnitude of the lattice translation to be evaluated in experimental images.

Since heterogeneous catalysis depends on surface chemistry, insights into the structure of supported catalytic materials is vital for understanding catalysis. We have developed a mass spectroscopic technique using very high angle annular dark field images in a scanning transmission electron microscope (STEM), that gives the number of atoms per cluster. We have also developed a robust interactive computer program to analyze these images rapidly. In this proceeding, we will present our method and results on analyzing ultra-small and dispersed Pt clusters on amorphous carbon, as well as our preliminary results of Pt on γ-A12 O3.

The ultra-small Pt clusters on the amorphous carbon Cu grid were made by a proprietary technique. Imaging for the STEM-based mass-spectroscopic technique was performed on a Field Emission Gun (FEG) VG HB601 STEM operated at l00kV. The absolute measured intensity from the clusters were converted to scattering cross-sections, which can then be converted to number of atoms: More details of this method can be found in Singhal, Yang and Gibson.

Introduction

The morphology of Si surface has much effect on physical and chemical properties of the surface. It is well known that the structure of Si surface can be modified by deposition of metals, and extensive studies have been carried out on the modified surface structures of Si. These studies were carried out on clean surfaces of Si in an ultra-high vacuum (UHV). However recently, it has been shown by Kamino et al.[l] that molten Au atoms remove a surface amorphous layer on the Si surface, making the surface clean even in a non ultra-high vacuum. The present study is an extension to the previous one (Kamino et al.) and Si surfaces modified by-molten Al atoms in a conventional vacuum were observed by means of TEM.

Experimental

The specimen used in this study was a mixture of pure Al particles and Si particles, the latter being prepared by crushing a Si single crystal into pieces.

INTRODUCTION

It has been reported by many authors that metastable β-Ti alloys such as Ti-Nb, Ti-V and Ti- Mo indicate the negative temperature dependence of electrical resistivity at low temperature since Ames et al., found the dependence on Ti-Nb alloys in 1954. Williams et al., suggested that the phenomenon resulted from formation of athermal co-phase crystals. In order to clarify the origin, electron diffraction pattern analyses were carried out using a specimen cooling stage and both of start and finish temperatures of formation of athermal co-phase crystals were decided. However, it is necessary to reveal the morphology of athermal ω-phase crystals to study the relationship between the electrical resistivity change and appearance of athermal ω-phase crystals. On the other hand, although electron diffraction pattern analyses are able to identify the atomic structures of crystals, which appear the diffraction spots, it is not so easy to reveal the detailed morphology of the crystals. Also, it has to consider effects of electron irradiation to transformations.

Bimetallic alloy particles supported on high-surface area carbon are widely used as electrode materials. Because of their stability, Pt-Sn alloys are of particulate interest for a variety of applications. Although the reasons for the superior properties of these catalysts are not fully understood, the dispersion, structure, and compositional homogeneity of the alloy clusters, are clearly important factors in determining their catalytic activity. In order to identify the composition and structure of Pt-Sn base nanoparticles, various characterization techniques, mainly X-ray diffraction (XRD) and transmission electron microscopy (TEM) [1], have been applied. However, very often conflicting conclusions regarding to the structure of Pt-Sn particles have been reported. The focus of the present work is to examine a commercially available carbon supported bimetallic Pt-Sn catalyst, with a nominal Pt:Sn wt. ratio of 2:1 (500N and 900N samples) and 3:1 (E270 sample), using high resolution electron microscopy (HREM), electron and X-ray diffraction techniques, and microchemical analysis by energy dispersive X-ray spectroscopy (EDS).

We are developing a new family of heterogeneous catalysts for hydrogenation catalysis. Catalyst synthesis is accomplished using colloidal polymerization chemistry which produce high surface area xerogel catalysts. These xerogels have been synthesized by one-step sol gel chemistry. These catalysts contain ruthenium and modifiers such as gold occluded or incorporated in a titanium oxide matrix. The materials, especially the modified systems exhibit favorable performance in microreactor evaluations for hydrogenation reactions and exhibit high activities. Nanostructural studies have revealed that the materials contain dispersed catalyst clusters which are desirable microstructures for the catalysis since the majority of the atoms are exposed to catalysis and are potentially active sites.

The composition and atomic structure of the xerogel catalysts containing ruthenium and other metals have been examined using our in-house developments of environmental high resolution electron microscopy (EHREM) the atomic scale [1-3] and low voltage high resolution SEM (LVSEM)[4] methods.

A solid-liquid interface in the Al–Si system has been observed at near-atomic resolution by in-situ heating experiments inside transmission electron microscopes. Chemical mapping was also attempt to detect distribution of constituent atoms near a solid–hquid interface.

Mixtures of Al particles and Si particles, the diameter of which ranged from 200nm to 800nm, were mounted on a specimen–heating holder developed by Kamino and Saka and examined in a Hitachi H–9000NAR and a Hitachi HF–2000 electron microscope, operated at accelerating voltages of 300 and 200kV, respectively. Heating these specimens above the melting point of pure Al inside the microscopes resulted in the melting of the Al particles. The liquid Al reacted with nearby Si, leading to the formation of Al–Si alloy phase. An interface between solid Si and Al–Si alloy hquid and also an interface between solid Al and Al–Si alloy liquid were observed. The dynamical behaviour was recorded continuously using TV–VTR systems. The distribution of Al and Si was visualized using a Gatan imaging filter GIF analyzer.

Small atomic clusters are of great importance for applications such as catalysts whose activity depends on the surface of the cluster. Attempts to determine the atomic short-range order and size of clusters have been made by analyzing the extended X-ray absorption fine structure (EXAFS). However, the analysis was made on an average of many small clusters. Analysis of extended energy-loss fine structure (EXELFS) in an electron energy-loss spectrum (EELS) has developed to the point where in some cases, the quality of the results is comparable to its X-ray analogue, EXAFS. No other technique provides nanometer-scale spatial resolution of the analyzed area for determining the atomic structure. Most EXELFS analysis has been performed on the K-ionization edge of lighter elements. For heavier elements, a more complex ionization edge such as the L-edge has to be used, due to the inefficiency of collecting high quality EEL spectra at higher energy-losses (Z > 18).

Nowadays there is great interest in the application of reflection high energy electron diffraction (RHEED) to monitor the growth of ultra thin films. This popularity of RHEED arose from the discovery of RHEED intensity oscillations in the early eighties. Namely, it was experimentally found that if the growth of a material follows a layer-by-layer mode then regular oscillating changes in the intensity of the specular beam occur, and the period of the oscillations corresponds to the deposition of one atomic layer of the material. These findings are of great practical importance and consequently RHEED is an important experimental technique used in nanoscale engineering. Nevertheless, the basic question of why RHEED oscillation are observed still remains open. In the past the Philips group suggested that intensity oscillations are a consequence of periodic changes in the roughness of the surface. About the same time, the University of Minnesota group claimed that during the growth the electron wave is reflected by two terraces (in an idealised case) and periodic changes of the interference conditions imply RHEED oscillation.

Since the original observations by Duncumb in 1962, a number of studies have been conducted on the effects of electron channel on characteristic x-ray emission and microanalysis. Most of the recent studies have concentrated upon using the phenomenon to perform site specific distributions of impurity elements in ordered compounds using the ALCHEMI methodology. Very few studies have attempted to accurately measure the effect as a function of orientation and compare these results to theory. In this study, two dimensional high angular resolution studies of channeling enhance x-ray emission were performed and herein the results are compared to theoretical calculations of Allen etal.

All experimental measurements presented here were conducted on a Philips EM 420T analytical electron microscope. The instrument was operated in the TEM mode, at 120 kV using an LaB6 electron source. The characteristic x-ray emission was measured using an EDAX ultra thin window Si(Li) detector having a FWHM of ∼ 145 eV at Mn Kα.

Introduction

We report here results additional to those presented earlier on single amorphous carbon nano-tips grown by carbon contamination on substrates of well-defined geometry in a scanning electron microscope (SEM). This method offers beneficial accuracy in controlling the properties of the nanoemitters such as length, aspect ratio and positioning on the given substrate.

The work function, emission field, current density, emitting area, solid angle of emission, energy distribution and short and long term stability of these emitters are all of interest. Since the literature does not provide, as far as we know, the work function of contamination-grown amorphous carbon, its value must be determined experimentally, in our case through Fowler-Nordheim (F-N) plots. This quantity and others can be derived once the electric field F at the emitting surface is known. Since the geometry of our emitters is reasonably well known from the fabrication procedure and subsequent electron microscopy, it is possible to find a satisfactory value for F by computation.

In 1970, Breedlove and Trammel predicted that, because of radiation damage, the direct imaging of individual organic molecules by electron or X-ray microscopy was not possible. This follows from an analysis, above IkeV, of the beam energy dependence of the ratio of inelastic scattering cross sections (some of which causes damage) to elastic cross sections (used to form the image). To minimize damage, we have therefore constructed a transmission electron microscope with nanometer resolution, which operates at 7 - 300 eV, below the threshold for most inelastic processes. Recently we have obtained our first images of TMV with this instrument. Isaacson first suggested that the K-shell ionization of carbon may be the most important damage process. Although less probable than valence excitation, it dumps more energy into the system, and the decay products are highly damaging. A subsequent study of electron beam damage in several π- bonded hydrocarbon crystals down to about 200 eV showed that a damage threshold does exist at about IkeV incident energy which can be explained by the carbon K-shell ionization. (Multiple scattering increases the energy threshold from 285eV).

The structure of small particles attracts great interest for their distinctiveness. Small particles of several materials undergo lattice parameter change or crystal structure change and sometimes form multiply twinned particles (MTPs) [1]. Configurational and structural fluctuations which they show under electron beam irradiation are also interesting phenomena [2]. This quasi-molten state are well explained theoretically [3], but the exact cause of it is still a matter of debate [2,4].

To clarify the behavior of nanoparticles under electron beam more precisely, we grew indium, palladium and molybdenum nanoparticles on Si (110) TEM samples in situ and investigated with an ultrahigh vacuum field emission transmission electron microscope (UHV-FE-TEM). It is a JEOL-2000VF operated at 200 kV with a base pressure of the column less thari 2.0 x 1O−8 Pa. Two kinds of evaporators are attached to the column in order to perform in situ evaporation of metals; an electron beam (EB) evaporator for high melting temperature material and a thermal heating evaporator with tungsten wire for low melting temperature ones.

Metallic nanoparticles finely dispersed onto high surface-area supports play an important role in heterogeneous catalysis. The performance of a supported metal catalyst can be directly related to the size and spatial distribution of the metallic nanoparticles. With the recent development of highresolution SEM instruments, it is now possible to observe nanoparticles in a field emission SEM. At low voltages, surface details of catalyst supports as well as metallic nanoparticles can be observed. The particle contrast in low voltage SEM images, however, is still not well understood. We have previously shown that the contrast of metallic particles can be enhanced if a small positive potential is applied to the sample. It is suggested that backscattered electrons (BE) significantly contribute to the visibility of metallic nanoparticles in high-resolution SE images. In this paper, we report further study on the origin of particle contrast in high-resolution SE images.

Figure 1 shows a set of SE images of the same area of a carbon supported Pt catalyst.

The development of the very large and ultra large scale integrated technology technology is based on the shrinkage of the lateral dimensions of semiconductors devices to a sub-micron level. Under these conditions the precise control of the total amount of dopant atoms dissolved in the silicon substrate, and of their depth and lateral distribution, become crucial requirements. One of the most promising device characterization technique is that based on transmission electron microscopy (TEM) analysis. Indeed, by means of TEM observations the device morphology can be imaged with a spatial resolution of the order of 0.1 — 0.3 nm; and, after an appropriate sample preparation method, based on selective chemical etch of doped silicon regions, the 2D dopant concentration profiles can be delineated with a resolution of the order of 5nm.

The chemical solution used for this aim consists of a mixture of both nitric and hydrofluoric acid.