Table of contents

A giant magneto-electric effect was observed in single-crystal samples of metastable orthorhombic ferro-electric ferro-elastic paramagnetic beta ' phases of rare earth molybdates Tb₂(MoO₄)₃ and TbGd(MoO₄)₃. The electric polarization in these compounds showed large and sharp changes in pulsed magnetic fields of 100 kOe at 78 K. Magnetic fields applied in a certain direction produced irreversible alterations in ferro-electric domain structure of these compounds that were observed visually. This effect was explained qualitatively in terms of large spin-orbit coupling in rare earth triply positive-charged ions and of interaction between the crystal field and the anisotropic charge density cloud of 4f electrons of rare earth triply charged ions of non-zero angular momentum. Related anomalies in the magnetostriction and in magnetization were also observed. The photoluminescence related to Tb³⁺ ions was observed at temperatures in the range 4.2-300 K. The most intense photoluminescence takes place in the green part of the spectrum. Here two groups of narrow sharp lines were observed, each group consisting of three lines. The photoluminescence exhibited a strong dependence of the polarization upon the ferro-electric domain orientation. Some applications of the effect are proposed.

Fe-SiO₂ and Fe₂₀Ni₈₀-SiO₂ granular solids have been prepared using a sol-gel method and investigated by X-ray diffractometry, Mossbauer spectroscopy and transmission electron microscopy. The magnetic properties were measured and the coercivity of these granular solid samples was found to be almost independent of temperature in the range 80-300 K. The effective magnetic anisotropy including shape anisotropy and strain anisotropy has been determined from the magnetizing curves of these granular solid samples using the law of approach to saturation. The obtained values of the effective magnetic anisotropy exceed 10⁶ erg cm^-3. The almost temperature-independent effective magnetic anisotropy can approximately explain the variation of H_c versus T for the granular solid samples. The shape anisotropy has been estimated assuming that the average axis ratio of the metal particles c/a is 1.2 and the strain anisotropy has been calculated from the effective anisotropy and the shape anisotropy. The magnitude of the strain anisotropy and the variation of the strain anisotropy with temperature are explained for the Fe-SiO₂ and Fe₂₀Ni₈₀-SiO₂ granular solid samples.

The photovoltage and photocurrent arising from CdTe/light yellow-S^2-, S₂^2-/C 'wet' solar cells has been measured as a function of various parameters. The effects of the S^2-/S₂^2- ion concentration and the dye concentration have been studied in detail. Increase of the S^2-/S₂^2- ion concentration improves the efficiency of the cell while it also increases the response time. 'Light yellow' dye is found to improve the stability of the cell.

This paper deals with the electromagnetic scattering of a plane wave obliquely incident on an infinite penetrable cylinder in uniform motion with velocity v along its axis. The analysis is carried out by means of a hybrid technique, which combines the finite-element method and the well-known modal expansion of a field scattered by cylindrical structures. The method presented allows for analysis of moving cylinders of arbitrary cross section and material properties.

S₁ to S₃ excited singlet state absorption and two-photon absorption in Rhodamine 6G at the pump wavelengths of 532 and 1064 nm respectively are investigated. The advantages of employing the pulsed photoacoustic technique for conveniently observing excited singlet state absorption are discussed. It is shown that, since photoacoustics and fluorescence are complementary phenomena, analysis using both techniques will yield a better understanding of optical processes in molecules like Rhodamine 6G.

Free oscillations of the keyhole in penetration laser beam welding are studied theoretically with regard to characteristic frequencies, damping rates and stability at large amplitudes. The normal modes form a discrete set which may be characterized by axial and azimuthal numbers. Due to viscous damping, only the lowest modes survive many oscillation periods, which yields a limited range of frequencies for the dynamic response of the keyhole to fluctuations of external welding parameters.

This paper presents a phenomenological analysis of the relaxation model for heat conduction and generation from the point of view of its consistency with the second law of thermodynamics. The formulation of the second law based on classical irreversible thermodynamics as well as that based on extended irreversible thermodynamics are considered. It is shown that introduction of the transient capacity of the heat source into the energy conservation equation does not lead to any inconsistency with the second law. An analogy between the heat transport and the motion of a particle in the resistive medium is presented. The analysis shows that an inherent property of Cattaneo's model for heat conduction is that heat may flow spontaneously from cold to hot regions.

A theoretical analysis is carried out on the transfer of heat into the workpiece in deep penetration welding. Special attention is given to the effect of melt flow on the transfer efficiency. It is found that the widely used model by Noller, which disregards the melt flow past the keyhole, overestimates the heat transfer. The difference is most significant at high welding speeds. The temperature distribution obtained in this work is more confined, leading to a shorter weld pool.

Gradual formation of a thin liquid film on a hot/cold rotating disc under the action of thermocapillary force is studied analytically. It is found that the thermocapillary force acting on the free surface enhances the thinning rate for cooling the disc.

Density fluctuations exist and propagate in two-phase flow as well as in fusion plasmas. Local flow parameters, such as velocity, mean density and the (directionally dependent) correlation length can serve to characterize a flow in a very detailed manner, including its topological structure (flow regime). Thus, determination of these parameters is an important task. This paper presents a non-intrusive correlation-based method for determination of local flow parameters. It is based on a model of the correlation structure of the flow. The method, and the model on which it is based, were developed earlier for two-phase flows with isotropic correlations. It is now extended to non-isotropic flows in which the correlation tensor can be diagonalized. With the present method it is possible to determine both axial and radial correlation lengths. Possible applications include diagnostics of non-isotropic two-phase flows as well as tokamak fusion plasmas.

A new model of CO vibrational kinetics is developed. The basis of the model is the multi-quantum vibrational exchange rate constants given by Billing (1986). The full Billing model of multi-quantum VV exchange gives rise to satisfactory agreement with experimental data on vibrational distribution functions in CO-N₂-He mixtures without needing any fitted parameters.

The measurements of the drift velocities and characteristic energies of electrons in deuterium are presented. Drift velocities are obtained between 3 and 125 Td by a standard Bradbury-Nielsen technique with a double set of shutters. The characteristic energy for the transverse diffusion was measured by a standard Townsend-Huxley technique to be between 3 and 1000 Td. The drift velocities and characteristic energies are in good agreement with the best available data in the range of overlap and extend the availability of the experimental data to higher E/N. The predictions based on the best available cross section sets do not match the experimental results at higher E/N above the limits of the previously available data.

Radiation emission and absorption in SF₆ arc plasmas are important energy transfer processes. Exact calculations, though possible in principle, are usually impossible in practice because of the need to treat about 400 spectral lines and also continuum radiation spanning wavelengths from 10 nm to 10 mu m. From calculated spectral absorptivities at 300000 radiation frequency points for SF₆ plasmas of 1, 5 and 10 bar from 300 K to 35000 K, we have calculated two integrals over radiation frequency. These integrals, designated Som and Delta Sim are used in the approximate method of 'partial characteristics', as formulated by Sevast'yanenko. The validity of this method of partial characteristics has been demonstrated by comparing exact calculations with the approximate calculations to evaluate radiation intensities, radiation fluxes and divergence of radiation fluxes for specified temperature profiles. Agreement to within 20% is obtained with exact calculations, but with a reduction of computation time of about four orders of magnitude. Furthermore, the new method has been used to evaluate net emission of radiation as a function of position in an algorithm to predict temperature profiles of wall stabilized arcs of any given current. Fair agreement is obtained between predicted and experimental values of central are temperatures and electric field strengths as a function of current for SF₆ arcs of radius 0.25 cm at a pressure of 1 bar.

The boundaries of the existence of a laser-sustained argon plasma in forced convective flow were investigated experimentally and numerically. The plasma was maintained by a cw CO₂ laser with an output power of 2.5 kW and burnt in free space at atmospheric pressure. Flow velocities changed from 2.3 m s^-1 to 8 m s^-1. The f number was 8.8. The results show that a quasi-2D model in which the axial flow is given by the relation rho u= rho ₀u₀( rho / rho ₀)¹2/, where rho ₀ and u₀ are the density and velocity of the cold gas respectively, describes the observed phenomena fairly well.

Oscillations of the electrode current, plasma density and plasma potential, caused by the potential relaxation instability in a weakly magnetized discharge plasma are studied. The oscillations appear when the electrode bias exceeds the plasma potential. As long as the difference between the electrode bias and the plasma potential is below approximately 2 V, that is approximately kT_e/e₀ in our experiment, the oscillations are periodic. When the electrode bias is increased above this value, the oscillations become irregular. The limit cycle in the phase space disappears, the power spectrum becomes broader and the autocorrelation time shorter. Presence of f^{- alpha} noise is observed in the power spectra. Distribution of Fourier phases for electrode current, plasma density and plasma potential oscillations is shown. Dimensional analysis of the signals is performed using the Grassberger-Procaccia algorithm. The correlation dimension is found to be between 1.2 and 2.0 for low electrode biases and no saturation of the In(C(r)) versus In(r/r₀) plots is found for higher electrode biases. Entropy K₂ is plotted versus r/r₀. The largest Lyapunov exponent is calculated using two different methods, one developed by Rosenstein and co-workers and the second by Wolf and co-workers. It appears that the largest Lyapunov exponent is positive and close to zero. An attempt to calculate the complete Lyapunov spectrum with the method of Kruel and co-workers is also made.

Anode and cathode potentials for electrodes of mild steel in stagnant air or flowing argon and CO₂ have been obtained both of the voltage distribution during arc quenching and of a measured melting velocity. The potentials are independent of arc current in the observed region 40-200 A. The energy for melting a unit length of a wire electrode has been established as the energy input for breaking a short-circuit bridge of the observed wire. The important role of neutral particle flux to melted electrodes has been revealed. A model of rapid arc quenching due to temperature and field emission of electrons has been suggested.

This paper extends earlier work by including ion diffusion in the transport processes for both positive and negative ions in a positive column where negative ions dominate over electrons and ion-ion recombination is the dominant recombination process. As anticipated in previous work, the effect is to preserve the structure of a central negative ion-positive ion plasma core surrounded by an electron-positive ion plasma. Increasing tau , the ratio of ion temperature relative to electron temperature, results in a progressive smoothing of the previously sharp boundary. Introduction of another parameter tau means that the problem is effectively a triple eigenvalue problem with I= beta _in_e0/ nu _i the ratio of central ion-ion recombination ratio to the ionization rate now being a function of tau . I is found to be a relatively weak function of tau , but the principal eigenvalue lambda is found to vary strongly with tau . An alternative method of approach using perturbation theory and tau as the variable for expansion is described and some results given to show consistency.

The dielectric barrier is becoming popular as a source of atmospheric pressure non-thermal plasmas. For atmospheric pressure plasma processing, either coaxial or parallel plate geometry can be used. In this paper, experimental investigation of the electrical characteristics of a coaxial dielectric barrier discharge is reported. The ratio of the inner to outer electrode radius has large influence on the corona onset voltage and the generation of radicals. It was also found that low levels of ultraviolet irradiation increase the number of microdischarges considerably resulting in a more uniform discharge. Also the energy coupled to the discharge from the power source increases with ultraviolet irradiation.

In this first part, the gaseous nature of filamentary positive streamers is demonstrated by the influence of hydrostatic pressure on their propagation in cyclohexane and pentane up to 7 MPa. At a fixed voltage, when the pressure is increased, the propagation velocity is found to be constant whereas the stopping length of streamers is greatly reduced. Correlated to this effect, the duration of transient currents and light emission signals is reduced. These effects are discussed in terms of the hypothesis that the conductivity of the streamer results from a gas discharge mechanism occurring within the filaments. These results constitute the experimental basis on which a study of the filament dynamics has been developed (part II).

This paper presents a study of the dynamics of positive streamer filaments in pentane using an optical method described in the text for pressures up to 9 MPa. It is observed that filaments expand and collapse, in a similar way to cavitation bubbles. Good agreement with the Rayleigh model shows that the dynamics of filaments is determined by liquid inertia. The influence of electric forces (electrostatic pressure) on the dynamics is found to be negligible. An evaluation of the energies involved shows that filaments are mainly composed of vapour, whose pressure varies with time and space and is on average higher than the hydrostatic pressure. Vaporization results mainly from energy dissipated at the filament extremity, which may be compared to a propagating point heat source of about 10 W power. At low pressure, a significant influence of energy dissipation within the filament, attributed to the transient current flowing during propagation, is also observed. At high pressure and/or high voltage, the stopping of the streamer is due to collapse of the gaseous filaments.

Electron energy distribution functions in glow discharges (I=5-50 mA, p=200-1000 Pa) were derived from the intensities of Ar and He spectral lines and N₂ molecular bands with different excitation thresholds. The relevant rate coefficients represent weighted integrals in the range from 12 to 24 eV. In molecular gases, structures in the low-energy region was inferred from theoretical calculations. An analytical formula with one experimentally adjusted parameter was used to describe the energy distribution. The total electron density was calculated from the electrical conductivity. The mean electron energies from three spectroscopic criteria, as well as those from an ionization-diffusion model of the discharge, are in close agreement for values of the fitting parameter characteristic for the particular plasma conditions. The high-energy part of the distribution functions is not far from a Druyvesteyn shape both in pure nitrogen and in noble gases. The electron densities in nitrogen and mixture plasmas were also measured from N₂⁺ molecular band intensities. Because of the rapid decrease in electron population at higher energies, the ionization of neutral N₂ into the upper N₂⁺ level can be ignored and these bands reflect the N₂⁺ ion ground state population. The excitation rate coefficients were calibrated in pure nitrogen and then applied to mixture plasmas, where the N₂⁺ ion fraction was calculated including charge exchange processes. Agreement with n_e results from electrical conductivity is very good over the entire parameter range investigated.

An experimental study of positive corona discharge in air for wire-to-plane geometry is presented for small gaps (from 8 to 16 mm). The measurements carried out show that wire-to-plane positive corona behaviour deviates from the Townsend relation in the low-current limit near the onset. This deviation is attributed to non-uniformity of the wire diameter and the roughness of the surface of the wire. We have introduced a correction to the Townsend relation based on a model of progressive onset potentials: the necessity of this correction is shown by means of an error analysis of the experimental data. The prefactors K₁⁺ obtained are of the same order of magnitude as those achieved for other reported geometries. Transverse current density distributions have also been measured and the experimental data compared with the Warburg law. A simple analytic model of the discharge is derived: the discharge is modelled by considering the drift of ions from the coronating wire electrode to the grounded plane as a succession of non-interacting cylinders of charge. The space charge deposited by the cylinders on their arrival at the plane gives a current distribution that fits well with the experimental data.

We have investigated the light switching properties of polymer dispersed liquid crystal (PDLC) films, which can be characterized by the spectral change in transmittance and reflectance due to switching by application of a suitable electrical voltage. The PDLC samples were prepared using the standard liquid crystal mixture E7 in a UV-curable matrix material. Using these samples, a transmission reduction of up to 30% was achieved in the solar spectral range. A strong dependence of the light switching behaviour on the mass ratio of liquid crystal and matrix material as well as on the curing rate of the matrix material was found, whereas the film thickness has only a small influence. Light scattering measurements allowed us to determine the average size of the scattering centres. In the samples investigated the liquid crystal droplets have radii below 2 mu m. Additional measurements of the electric properties of the PDLC films showed that the capacity and the resistivity of the PDLC films depend on both the applied AC voltage and its frequency. The power consumption of the above PDLC films is in the range of 5 W m^-2, if powered with a line frequency of 50 Hz; at lower frequencies (e.g. 20 Hz) the power consumption can be reduced to about 2.5 W m^-2 without lowering the change in transmittance significantly.

A simple quasi one-dimensional self-consistent model of the keyhole in penetration laser welding has been given previously. By considering the equation of heat conduction and the pressure balance at the keyhole wall, the radius of the keyhole was obtained. The keyhole was assumed to be a circle, concentric to the laser beam. In the present paper we drop these restrictions. The keyhole boundary is considered to be a free curve, adjusting itself to the physical mechanism previously formulated. The so obtained Stefan problem is solved numerically with a finite element method. The keyhole boundary deviated only slightly from a circle but was shifted with respect to the laser beam, depending on welding speed and laser power. The temperature is found to be nearly constant at the keyhole boundary. Our results give justification to previous investigations, which considered a constant temperature along the keyhole wall as a boundary condition for the equation of heat conduction.

Diffusion-induced stresses in a long bar with a square cross section under an electric field are calculated. We obtain in advance analytical solutions of the concentration, then with these results, the diffusion-induced stresses are also calculated analytically by introducing the displacement potential and Airy stress function. The results show that the electric field can depress the stresses and deviate the positions of local maximum stresses, and these effects are more apparent at short times than at long times.

Thermal simulations made on ceramic substrates have been compared with experimental thermographic measurements. Very good agreement has been obtained. The most difficult point turns out to be determination of the convection coefficient.

Laser-ablation deposition of films with nominal composition Fe₈₁B_13.5Si_3.5C₂ from targets of Metglass 2605SC ribbons onto room temperature substrates in vacuum is reported. It is found that the density of droplets on the surface of the film is controlled mainly by variation of the laser pulse energy. Scanning electron microscopy and energy-dispersive x-ray analyses indicate that the films are ablated as stoichiometric phases. X-ray diffraction and conversion electron Mossbauer spectra show a line broadening of the amorphous pattern compatible with a decrease in average grain size compared with the ribbon target. Analysis of the hyperfine field distribution at Fe sites indicates a 23% content in nano-sized particles in the ablated films, with the particles having an estimated average grain size of about 15 nm.

Electrical conductivity, mobility, cationic and anionic transference numbers and numbers of respective mobile charge carriers have been measured for polyethylene-oxide-based polymer electrolytes complexed with mixed anion salts. The systems studied are (polyethylene oxide)₁₅: ((1-x)NaI+xNaSCN) and (polyethylene oxide)₁₅: ((1-x)NaI+xNaClO₄). It is shown that the presence of mixed anions in the polymer salt complex electrolytes changes the relative number of mobile cations and anions together with their respective transference numbers as well as the crystallinity of the polymer matrix. A qualitative correlation has been established between the crystallinity, t_cation, conductivity and the number of mobile anions/cations. The maximum change in the cationic (Na⁺) transference number was for polyethylene oxide: NaI for which t_Na⁺ could be increased from 0.18 to about 0.6 in the mixed anion system (polyethylene oxide)₁₅: (0.35NaI+0.65NaSCN) with comparable conductivity.

Using photolithographic techniques, four-probe electrical resistivity measurements of Nb-doped BaTiO₃ ceramics, at the microscopic scale, are presented from room temperature up to 180 degrees C. The PTC effect displayed by several grains exhibits an abrupt jump at the Curie temperature, followed by a monotonic quasi linear increase of the resistance. This behaviour looks quite different from the usual PTC effect observed on macroscopic ceramics. The jump is clearly due to a change of resistance within the bulk. It appears, from the microscopical results, that the bulk itself must be partially involved in the increase of resistance above T_c contrary to the generally accepted theory.

Shallow traps for p-type charge carriers were found in a ladder-type poly(p-phenylene). Using the thermally stimulated current technique a trap depth of 0.1 eV was determined. The electrical current of a gold-ladder-polymer-gold structure at field strengths of 4*10⁴ V cm^-1 is space-charge limited. We propose a Gaussian trap distribution centred at 0.1 eV above the valence band edge.

A vapour phase epitaxial layer of ZnTe was grown on (111), (110) and (100) planes of In-doped n-CdTe substrates. The I-V characteristics of the CdTe-ZnTe heterojunction show that the generation and recombination currents and the diffusion current are controlled by the interface states of the junction. The frequency-dependence and temperature-dependence of the C-V characteristics of the heterojunction were studied. The results show that the band bending is 1.06 eV. The length of the transition region is found to be 0.004 mu m toward the n-CdTe side and 4.81 mu m toward the ZnTe side, which means that the situation is as if an n⁺-p junction had been formed. The interface state density has been found to be 7.5*10¹³ cm^-3. An attempt was made to draw the complete band diagram of the heterojunction.

Single-frequency measurements of dielectric relaxation parameters at different concentrations are used to estimate the double-relaxation times tau ₁ (smaller) and tau ₂ (larger) of some mono-substituted anilines in benzene at 35 degrees C for 2.02, 3.86 and 22.06 GHz electric fields respectively. The o- and m-anisidines like p-toluidines exhibit double-relaxation phenomena at 3.86 and 22.06 GHz whereas o- and m-toluidines show the same effect at 2.02 and 3.86 GHz respectively. Only p-anisidine, however, shows the mono-relaxation behaviour at all frequencies. The relative contributions C₁ and C₂ towards dielectric relaxation for tau ₁ and tau ₂ are computed from Frohlich's equations only for comparison with those of the graphical techniques adopted here. The dipole moments mu ₁ and mu ₂ in terms of tau ₁ and tau ₂ are then determined from the slope beta of concentration variation of the ultra-high-frequency conductivity K_ij for those compounds to establish their conformations.

An exceptionally sensitive calorimetric technique has been devised for measuring microwave absorption by small samples of ultra-low-loss dielectrics at cryogenic temperatures. The loss factor of sapphire at 2.45 GHz diminished progressively with decreasing temperature from 3.5*10^-5 at 300 K down to 2*10^-10 at 4.2 K. The rate of decrease showed no tendency to attenuate at temperatures down to 4.2 K.

M₅(PO₄)₃X apatites are well known for their technological importance as phosphors, laser hosts and biocompatible materials. Divalent-europium-activated alkaline earth chloroapatites are of special importance for their application as the blue component in high-efficiency trichromatic fluorescent lamps. In these apatites, Eu²⁺ yields narrow-band emission in the blue region corresponding to the 4f⁶5d to ⁸S_7/2 allowed electric dipole transition. It has been found that the dominant emission band observed can be assigned to Eu²⁺ occupying M_II sites of the apatite system having C_1h symmetry. In the case of barium chloroapatites one could also observe Eu²⁺ emission from a second type of site (M_I) that is available in the apatite system. The weak emission observed in the latter can be attributed to the intense spectral overlap with the former site(s) and thermal quenching effects. The exchange coupling between the 4f⁶ and 5d electrons of Eu²⁺ manifests itself in 'stair-case' features in the excitation spectrum and the strength of the exchange interaction is highly dependent on the host apatite(s). Also, in the Eu²⁺, Mn²⁺ co-doped system, energy transfer is found to occur from the former to the latter and the transfer is regulated by an exchange process. Various results based on these facts are discussed in detail.

The thermal sampling technique is applied to investigate relaxation processes in amorphous blends and it was found that a distribution process exists. The relaxation kinetics are interpreted in terms of rate theory. A linear relation between activation enthalpy and entropy values is obtained, indicating that the compensation law is valid. The compensation plot revealed that the energy of the blend relaxation process comprises not only the energy required to move the segments along the reaction path but also the energy required to create these segments.

Synthetic diamond single crystals synthesized with FeCo as solvent/catalyst were heated at 1100 and 1650 degrees C for up to 1 h in vacuum. X-ray diffraction photographs taken at room temperature showed the presence of (i) body-centred cubic randomly oriented FeCo inclusions, (ii) slightly textured graphite and (iii) slightly textured compressed graphite. The FeCo inclusions remained stress-free even in crystals with the maximum, 30-35 kbar, of the observed residual graphite compression. It is concluded that the compressed graphite is located in cracks extending from the FeCo inclusions and that it is the cause for failure of synthetic diamonds on heating.

An experimental investigation has been carried out to study variations of microstructure, composition and electrical properties (sheet resistance and its temperature coefficient) occurring in thick film (cermet) resistors during the annealing process which transforms the initial ink into a resistive layer. Resistors based on Bi₂Ru₂O₇ and a lead silica glass have been studied. Analyses have been performed on the inks using thermogravimetric measurements, and on annealed (fired) layers by means of X-ray diffraction, atomic absorption, electron microscopy, microprobe analysis and other complementary techniques. The results point out the role of exchange reactions and redox reactions inside the resistor body and emphasise the complexity of the phenomena which concur to define the final electrical properties of these resistive systems. Interrelations between microstructure, composition and electrical properties have been found and tentative explanations proposed.

The applicability of a near-field scanning optical microscope (NSOM) to measure ultrafast phenomena is experimentally demonstrated. For this application a commercial NSOM is combined with a picosecond laser sampling system. The ability of this time resolved NSOM to perform function and failure analysis of microwave devices is demonstrated. Measurements of high speed electric signals up to 7.5 GHz via the direct electro-optic sampling technique are performed.