Search

In principle, electron microscopy can provide accurate and precise numbers for the unknown structure parameters of materials under study. As an example, atomic column positions can be determined with accuracy and precision that is orders of magnitude better than the resolution of the microscope. This requires, however, a quantitative, model-based method. In our opinion, the maximum likelihood (ML) method is the most appropriate one since it has optimal statistical properties. This contribution aims to explain the basic principles and limitations of the method. Moreover its practical applicability and usefulness will be demonstrated using recent experimental examples.

In graph mining, a frequency measure is anti-monotonic if the frequency of a pattern never exceeds the frequency of a subpattern. The efficiency and correctness of most graph pattern miners relies critically on this property. We study the case where the dataset is a single graph. Vanetik, Gudes and Shimony already gave sufficient and necessary conditions for anti-monotonicity of measures depending only on the edge-overlaps between the intances of the pattern in a labeled graph. We extend these results to homomorphisms, isomorphisms and homeomorphisms on both labeled and unlabeled, directed and undirected graphs, for vertex and edge overlap. We show a set of reductions between the different morphisms that preserve overlap. We also prove that the popular maximum independent set measure assigns the minimal possible meaningful frequency, introduce a new measure based on the minimum clique partition that assigns the maximum possible meaningful frequency and introduce a new measure sandwiched between the former two based on the poly-time computable Lov\'asz $\theta$-function. acceptance rate = 9.7% ispartof: pages:73-82 ispartof: Proceedings of the IEEE International Conference on Data Mining pages:73-82 ispartof: IEEE International Conference on Data Mining location:Pisa, Italy date:15 Dec - 19 Dec 2008 status: published

In this abstract it will be shown that exit waves are very useful in order to obtain quantitative information on the projected atom column positions and on the composition of these columns. According to the channelling theory [1], an atom column acts as a channel for the incoming electrons in which the electrons scatter dynamically without leaving the column. This theory gives a simple mathematical expression for the complex exit wave including only the lowest bound state of the column potential, which is called the 1s state. Its amplitude is peaked at the atom column positions whereas the phase, which is a constant over the column, is proportional to the average mass density of the column [2]. A convenient way to visualize these effects is by plotting the complex value of the exit wave for each pixel located at a projected atom column position in a so-called Argand plot with the x- and y-coordinate corresponding to the real and imaginary value, respectively [2, 3]. It turns out to be more convenient to shift the exit wave over a vector corresponding with the entrance wave so as to yield a fingerprint of the electron interactions only. The column position is chosen since its value is most sensitive to absorption, the Debye-Waller factor and the scattering factors. For a wedge-shape structure containing different thicknesses, the Argand plot is expected to be a circular locus which starts at the origin (for zero thickness) and which has its center on the x-axis. The radius of the circle is then the amplitude of the scattering wave, which should be identical for different atom column types if the columns are well-separated. This can also be proved using an empirical scaling behaviour which is valid for all the 1s states [4].

Transmission electron microscopy (TEM) with aberration correction provides one of the most powerful techniques for revealing atomic scale structure of solid samples. This technique has been widely used to solve the structure of crystalline samples. In practice, a lot of materials that are of key applications are either highly disordered or amorphous. The fact that the many interesting nanostructures are three dimensional and disordered makes them at the same time also much more difficult to study. Indeed in such structures many atoms are at non-crystallographic equilibrium positions and are thus much more prone to displacive radiation damage. This is essentially true in electron tomography [1], where the recording time is large, the accumulation of these defects during the tomographic acquisition process can lead to a significant degradation of the object, and hence to a significant influence of irradiation on the object reconstruction. Moreover, due to the improvement of resolution of TEM, the flux of electrons that will irradiate a local region in the experiment is rather large. Radiation damage is a serious problem also for structure analysis in nanomaterials that have many atoms on the surfaces, which tends to migrate under the illumination of electrons. This poses a fundamental limit on the ultimate resolution of the technique; especially for amorphous structures and it is questioned whether atomic resolution tomography is possible at all for such structures. Developing TEM technique for solving the structure of amorphous materials or highly disordered is of great challenge to today’s microscopy.

Inelastic holography is a technique in which electron energy loss spectroscopy (EELS) and electron holography are combined to study coherence effects in inelastic scattering. In typical inelastic cross section calculations it is assumed that we have one scattering atom and a single incoming plane wave. Although this setup seems to provide the answer to most of the practical inelastic scattering problems, there are several cases where it fails. The reason for this failure is the existence of interference effects in the scattered wave. This occurs when there is more than one incoming plane wave and/or when there is more than one scattering atom; like in almost every experiment in the electron microscope.

A non-destructive method for measuring the thickness of thin amorphous films composed of light elements has been developed. The method employs the statistics of the phase of the electron exit wave function. The accuracy of this method has been checked numerically by the multislice method and compared with that based on the mean inner potential.

The relation between the contrast in an HREM image of a substitutional alloy and the local composition is studied within the framework of the real-space theory of electron scattering. Two algorithms are derived to determine the composition: a deconvolution and a convolution. The algorithms are tested on simulated images of Al x Ga 1− x As, of a GaAs/AlAs interface and of an Al x Ga 1− x As alloy with randomly occupied group III sites. The deconvolution method is very sensitive to small deviations from the linear relationship between image contrast and composition, and to the influence of neighbouring columns. The more robust convolution method works best for defocus values optimizing the differences in contrast between GaAs and AlAs images, and for sample thickness around 10 nm. It is estimated that under optimal imaging conditions compositions can be determined within separate columns of group III atoms with an accuracy of ±0.10.

The propagation of elastic and inelastic electron waves in high-energy electron diffraction in a thin crystal foil is described in a real-space framework. Within the usual approximations of the “elastic” real-space method and applying the “single-elastic-into-inelastic” approach, a real-space slice expansion for inelastic electron waves is set up, which can be used for any type of inelastic scattering. Single-electron-excitation inelastic processes are discussed in more detail. In real space, the imaginary absorption potential for single-electron excitation can be computed for each atom individually, independent of the crystal structure and the accelerating voltage. In this work, these absorption potentials have been computed explicitly for all types of atoms and parametrized by a linear combination of four Gaussians. These parameters are tabulated.

The induction of drug-metabolizing enzymes in rat liver was studied after subchronic administration of the new triazole antifungal agent fluconazole. The administered doses were 10, 40, and 160 mg/kg per day for 7 days. Fluconazole behaved as a high-magnitude inducer and significantly increased cytochrome P-450 concentrations already at 10 mg/kg (+42%). Cytochrome P-450 induction by fluconazole was dose dependent and reached a value of 302% of the control value at the dose of 160 mg/kg. The induction effects on cytochrome P-450 were also reflected in the drug-metabolizing enzyme activities in hepatic microsomes of pretreated rats. Fluconazole (160 mg/kg per day) preferentially induced the demethylase activities of N,N-dimethylaniline and p-nitroanisole to 258 and 281% of the control values, respectively. The detoxification enzyme UDP-glucuronosyltransferase was significantly lowered by fluconazole at the highest dose. A possible link between the induction potential and the pharmacokinetic properties of triazole antifungal agents is discussed.

We are living in a vely exciting period for structural research using HREM. Indeed, the possibility to ‘see the individual atoms of which matter is constituted seems within reach. Recent technological improvements allow to obtain a resolution of about 0.1 nm. However, the potential power of the technique is still severely limited by the problem of quantitative interpretation of the images. For instance, the use of computer simulation images requires much a priori knowledge which makes HREM very dependent on other techniques. The situation can be compared with the early days of X-ray diffraction. Recent developments make it possible to retrieve the object structure directly from the electron micrographs. We will discuss future prospects in this direction and we will show that the ultimate resolution is finally limited by the object itself.

The three-dimensional (3-D) shape and density distribution of a cod otolith was determined using a newly developed micro X-ray tomograph. The technique requires no special treatment of the otolith and could yield the entire 3-D growth history in a single analysis. 999 The Fisheries Society of the British Isles

It is proven in a rigorous way that the frozen phonon model that is currently used to account for phonon scattering in multislice calculations, is equivalent to a full quantum-mechanical treatment of the inelastic phonon scattering process.

X-ray microtomography (micro-CT) is a new technique allowing for visualization of the internal structure of opaque specimens with a quasi-histological quality. Among multiple potential applications, the use of this technique in otology is very promising. Micro-CT appears to be ideally suited for in vitro visualization of the inner ear tissues as well as for evaluation of the electrode damage and/or surgical insertion trauma during implantation of the cochlear implant electrodes. This technique can greatly aid in design and development of new cochlear implant electrodes and is applicable for temporal bone studies. The main advantage of micro-CT is the practically artefact-free preparation of the samples and the possibility of evaluation of the interesting parameters along the whole insertion depth of the electrode. This paper presents the results of the first application of micro-CT for visualization of the inner ear structures in human temporal bones and for evaluation of the surgical positioning of the cochlear implant electrodes relative to the intracochlear soft tissues.