Table of contents

Calculations are presented for the elastic and inelastic image profiles for the transmission electron microscopy of amorphous (biological) materials, corresponding to objective lens defocus values, 0 to -2000 nm for an incident electron energy of 100 keV, used in the electron microscopy of biological specimens. These calculations are relevant to image analysis procedures which neglect the contribution of the inelastic electron scattering to the image intensity for these large underfocus values. For small defocus values, 0 to -200 nm, the inelastic contribution can be described as an unstructured background in the electron micrograph. However, corresponding to the larger underfocus values, -500 to -2000 nm, the elastic image resolution is inferior to that of the inelastic image. The inelastic image resolution is less dependent on the objective defocus value than the elastic image resolution, and in dark-field microscopy at defocus values -500 to -2000 nm, the inelastic image is of primary importance. In bright-field microscopy the dominance of the high-resolution elastic component is evident only for small underfocus values, but at the larger underfocus values the inelastic image is relevant.

A method is given for the evaluation, in transmission electron microscopy, of the amplitude and phase from the intensity distribution of an electron micrograph. The method requires a minimum of two micrographs taken under different defocus conditions. The iterative scheme requires only the relative defocus between micrographs, and the procedure is valid both in bright-field and dark-field microscopy for any specified coherence of the electron source. Assumptions on the scattering properties of the specimen, such as the weak-phase-weak-amplitude object, are not required. For a complete determination of the amplitude-phase distribution for electron transmission through the specimen, the electron micrograph must be corrected for the effect of lens aberrations and defocusing to give the electron wavefunction immediately after transmission; only in the case of a weak-phase object can this wavefunction be directly related to the projected potential distribution in the object.

Inelastic electron scattering is explicitly omitted from the analysis presented.

Focusing of EM beams in elemental semiconductors with energy-independent isotropic effective mass through the application of a radially inhomogeneous magnetic field has been proposed. For the typical case of a germanium sample, the periodic focusing length is found to be about 4·67 cm.

Measurements of the three unclamped principal-axis electro-optic coefficients of meta-nitroaniline are reported. One of these, r₃₃, is the largest yet found in a purely nonferroelectric crystal. Some consideration is given to the electro-optic modulator configurations which could operate profitably using this single coefficient.

Experiments have been performed in which the light output of an argonion laser was focused onto a thin film of gold deposited on a glass substrate. In addition to evaporation of the gold film, damage to the substrate surface has been observed for power levels as low as 0·1 W and for intensities of less than 5 kW cm⁻². The nature and extent of the damage have been found to vary dramatically with the illumination conditions and film thickness.

A quantitative procedure for the evaluation of object mass-thickness in electron microscopy by the use of a new recording method is described on the basis of single and plural scattering theories. With this method, recorded images can be observed as colour images, and the difference in mass-thickness of a specimen is estimated directly from the observed colours as well as from the photographic density. Because of the high contrast, the detectable limit of difference in thickness is found to be about ten times smaller than that with usual emulsions.

A new method for x-ray determination of the volume fraction of phases in a sample has been developed. This method is characterized by the accurate and rapid determination of these phases by analysing a single reflection from each phase, regardless of the condition of the sample.

This new method was applied to the determination of the quantity of retained austenite in some quenched carbon steels; it was found that the quantity of retained austenite increased with quenching temperature, becoming almost constant, and depended on the carbon content of the matrix and the quenching procedure.

An apparatus is described which is suitable for measuring the flexural damping and dynamic Young's modulus of beam specimens at stress levels up to 200 MN m⁻². Symmetric free-free flexural modes of vibration are excited by a coil/magnet pair, the coil being attached to the midpoint of the beam. Extraneous sources of damping have been reduced to a low level despite the difficulties of producing high-amplitude vibration. Low-damping specimens must be tested in vacuo (0·5 Torr was usually sufficient) if the effects of air damping are not to be significant. Specific damping capacities from 0·1 to 50% can be accurately measured over a frequency range of 100 to 800 Hz and a temperature range -50 to +200°C.

Fracture toughness tests have been conducted on polyester resin casts and laminates containing a resin flexibilizing additive. In spite of a large change in the resin strain to failure there was little change in the critical stress intensity factor of either the resin casts or laminates. Crack growth rates for the resin casts under cyclic loading could be represented in terms of stress intensity factor range ΔK through the power law relationship da/dN=A(ΔK)^m, where A and m are constants; m was approximately 5 for all resin flexibilities. The crack growth rate was constant provided that general yielding did not occur. Fracture surfaces were examined by scanning electron microscopy, where distinct ridge structures were observed in the crack growth specimens.

Results are presented which demonstrate the sensitive dependence of the onset time t_d for deviations from ohmic current values in thin acoustoelectrically active photoconducting CdS platelets, on the optical absorption of the photoexciting illumination used.

Computations, based on a theoretical model of linear round-trip acoustoelectric gain in thin platelets having exponential conductivity profiles parallel to the current flow, give semiquantitive agreement with experimental curves obtained using narrow-band illumination with measured optical absorbance values. A number of experimental and theoretical reasons are provided in explanation for the lack of precise agreement.

By the use of very weakly absorbed illumination and carefully prepared specimen contacts, excellent agreement is obtained between the experimental dependence on conductivity of threshold and cut-off voltages for acoustoelectric activity, and the predictions of the linear gain model with drift mobility as a `fitting' parameter. By the assumption of a reasonable value for the non-electronic acoustic loss, a reasonable fit is also obtained between experimental and theoretically predicted values of the voltage for minimum onset time.

A simplified approach of the Halperin and Lax model of the density-of-states tails in heavily doped semiconductors is presented: the trial wavefunctions are assumed to be hydrogen-like ones. Analytical expressions of the density-of-states tails are obtained, in very good agreement with the computed Halperin and Lax results.

The connection of the density-of-states tail with the unperturbed parabolic band is then achieved in the same way as previously done by Hwang, and numerical values are calculated for various doping levels in GaSb. The influence of the band tails on the electrical I-V characteristic, and on the optical luminescence spectra, is calculated. Good agreement is shown with experimental results on GaSb p-n junctions. Thus valuable information about the location and spectral properties of the recombinations in p-n junctions can be deduced from this simplified band-tail model.

Reversible memory switching in evaporated SiO films with one Co and one Ag electrode was investigated. The high-impedance state was more than 10⁹ Ω; the low-impedance state about 300 Ω. A rectified sine wave was used as a switching pulse in one direction and a short square-wave pulse in the other. The switching characteristics at different repetition rates and the properties of the low-impedance state suggest metal diffusion from the electrodes. A model for explaining the memory effect is presented.

An account is given of a method of measuring the yield and statistical distribution of the number of secondary electrons emitted from films of certain alkali halides under soft x-ray bombardment. The results are interpreted in terms of a theoretical model which is partly based on a Monte Carlo calculation used previously by the authors to describe the passage of low-energy electrons through thin films. Numerical values for parameters characterizing the electron-emission process are compared with results of other workers.

A segmented model domain wall structure is discussed which enables an accurate theoretical analysis to be made of the motion of domain walls severely distorted by an inhomogeneous eddy current field. Methods are developed for determining the wall profile and eddy current field at successive phases of the motion, when either the applied field or the flux is a prescribed function of time. These techniques are applied to a detailed study of an isolated domain wall driven so as to generate sinusoidal flux variations. Severe wall bowing is found to reduce the relative eddy current drag very roughly as the cube root of the frequency and the square root of the amplitude. At low flux amplitudes, asymptotic agreement is found with an independent exact theory of the infinitesimal motion of a continuous quasi-plane wall. The segmented model, adapted to include interactions between walls by eddy current overlap and actual merging of neighbouring walls into cylindrical domains, is also applied to the study of saturation sinusoidal induction.

Comparison with experiment is made by simulating the motion of a pair of domain walls observed by Helmiss (1969) in a detailed study of flux reversal in a single-crystal `picture- frame' specimen of 3·5% silicon-iron. Good agreement is found between the simulated eddy current losses and those reported by Helmiss. Reasonable concordance of calculated values of the surface wall velocity with those observed by Kerr magneto-optic methods is also obtained.

The total wall energy has been taken as the sum of exchange and anisotropy terms, and the Euler equation derived for the magnetization distribution through a wall. This equation has been solved numerically and the corresponding minimum wall energy found to increase monotonically with the mean radius of the wall. The upper bounds to the energy have been determined analytically and they show many features of the numerical calculations.

When LiF crystals containing a low concentration of Mg²⁺ ions are irradiated with γ-rays and then heated, five glow peaks may be observed. The intensity of dielectric absorption is used as a monitor of the state of aggregation of the Mg²⁺-vacancy pairs. It is shown that one of the factors determining the heights of peaks 2 and 3 is the concentration of dipoles in the crystal at the time of irradiation; those Mg²⁺-vacancy pairs present in trimers and larger aggregates make no contribution to the height of these peaks. Both dipoles and trimers contribute to the traps responsible for peaks 4 and 5, but aggregates containing more than three Mg²⁺-vacancy pairs do not contribute.

Previous work based on the assumption of a direct proportionality between dipole concentration and the height of peak 2 is invalid.

An expression is derived for the motionally averaged NMR fourth moment of a partially oriented polymer. This result, taken together with expressions previously obtained for the second moment, makes it possible to study quantitatively the effects of different types of molecular motion. The case of drawn polyoxymethylene is used to show how the theory may be applied. Detailed comparisons are made with the previous experimental results of Olf and Peterlin. For this system it is found that the experimental data at room temperature may be accounted for by a model in which the chains are oscillating about the helical axis with an amplitude of 42°. At a higher temperature (100°C) it is necessary to assume, in addition, that the chains are undergoing rapid translational motion which has the effect of greatly reducing the intermolecular contributions to the moments.