Table of contents

In modelling the plasma kinetics in dielectric barrier discharges (DBDs), the electron energy conservation equation is often included in the rate equation analysis (rather than utilizing the local-field approximation) with the assumption that the electron energy distribution function (EEDF) has a Maxwellian profile. We show that adopting a Maxwellian EEDF leads to a serious overestimate of the calculated ionization/excitation rate coefficients and the electron mobility for typical plasma conditions in a xenon DBD. Alternative EEDF profiles are trialed (Druyvesteyn, bi-Maxwellian and bi-Druyvesteyn) and benchmarked against EEDFs obtained from solving the steady-state Boltzmann equation. A bi-Druyvesteyn EEDF is shown to be more inherently accurate for modelling simulations of xenon DBDs.

The aim of this work is to study the properties of a silicon-based deposit realized with a glow dielectric barrier discharge at atmospheric pressure in a nitrogen, silane and nitrous oxide mixture. It is shown that a continuous thin film can be realized. The chemical composition of the thin layer has been determined by x-ray photoelectron spectroscopy and static secondary ion mass spectrometry. The characteristics of the deposit are correlated to those of the discharge. The first steps of a chemical pathway leading to the precursors of the deposit are proposed from the analysis of the optical emission spectrum of the discharge. It appears that, unlike the low-pressure PECVD in a N₂O-SiH₄ mixture, in an atmospheric-pressure glow discharge NO is the main oxidizing species, due to the action of the metastable nitrogen molecules.

We present a completely deterministic model for the electrical-tree breakdown mechanism in solids, and show that it is capable of generating branched structures even when the polymer is assumed to be homogeneous. The model is non-linear in the local electric fields and contains feedback features. These generate deterministic chaos in the discharges driving the breakdown, such as have been observed in laboratory experiments. It is surmised that it is this chaotic behaviour which is responsible for the branching.

The effect of a horizontal magnetic field of 333 mT flux density on the co-deposition of Ni and Fe from a modified Watt's bath was studied. Magnetic field imposition increases the iron content of the deposit and destabilizes cell potential variations under galvanostatic deposition.

We present measurements of the nanoscale elastic and viscoelastic properties of samples of poly(methylmetacrylate) (PMMA)/rubber nanocomposites. For these studies we have used a new technique based on atomic force microscopy (AFM) with ultrasonic excitation, heterodyne force microscopy (HFM), which provides a means of testing the viscoelastic response of polymeric materials locally (in tip-probed regions) at MHz frequencies. Phase-HFM contrast distinguishes local differences in the dynamic response of PMMA/rubber composites. Comparison of HFM with other AFM-based techniques (ultrasonic force microscopy, friction force microscopy and force modulation microscopy), while imaging the same surface region, emphasizes the unique capabilities of HFM for these kinds of studies, and reveals key nanostructural characteristics of the composites. Some of the toughening particles appear to be broken down, with areas of PMMA detached from the surrounding matrix.

Thin films of Te_48-yAs_30+yGe₁₀Si₁₂ (y = 0,5,10,15) of different thicknesses are deposited on glass substrates by thermal evaporation. X-ray diffraction revealed the amorphous nature of both powder and thin-film forms. The current-voltage characteristics in the temperature range 295-358 K and thickness range 75-300 nm are ohmic in the lower field regime followed by non-ohmic behaviour in the high-field regime, which has been analysed by the anomalous Poole-Frenkel effect. The temperature dependence of ohmic current is that of a thermally activated process with an activation energy increasing with increasing As content. The effect of composition and temperature on the dielectric constant have been studied. The dielectric constant increases with increasing As content.

The surface and interface roughness and roughness exponent of PbZr_0.53Ti_0.47O₃ (PZT)/La_1.85Sr_0.15CuO₄ bilayers deposited on SrTiO₃ (001) substrates by rf/dc magnetron sputtering have been measured by x-ray reflectivity and diffuse-scattering methods. We have found that the surface roughness increases and the roughness exponent decreases with the increase of the thickness of the PZT layers; and that there exist non-designed cap layers on the upper surfaces of the PZT layers. The growth character of the bilayer films is discussed.

The photoluminescence (PL) of the sol-gel derived Ce³⁺-doped silica (with or without Al co-doping) annealed at various temperatures was monitored. It was found that the PL shows a variation in emission intensity (first a decrease and then an increase), band position (red shift) and band width (broadening) with an increase of the annealing temperature. In other words, solution-like PL is exhibited in low-temperature-annealed samples and, with increasing the temperature, the PL becomes more and more similar to that observed in Ce³⁺-doped silica prepared by chemical vapour deposition. These changes can be attributed to the annealing induced structural evolution in silica: Ce³⁺ ions change from coordinating with water and silanols to embedding in silica network. In addition, Al co-doping markedly enhances the emission intensity and also affects the band position (blue shift) and band width (narrowing), which can be due to the changes in the local environment induced by the addition of Al.

The Hall-type ion source operated with nitrogen has been used in two ways in the synthesis of carbon nitride films: by ion beam nitridation of a graphite target and by simultaneous deposition on a collector (silicon wafer and nickel foil) of sputtered carbon atoms and nitrogen atoms from the surrounding nitrogen plasma. We present Rutherford backscattering spectroscopy, x-ray photoelectron spectroscopy and x-ray excited Auger electron spectroscopy results on the carbon and nitrogen atom fraction and chemical bonding state in carbon nitride films. The C 1s XPS peak of a sample deposited on silicon wafer was deconvoluted into three peaks at 284.7, 286.7 and 288.8 eV. The peak at 286.7 eV by analogy with polymethacrylonitrile (C≡N, 286.74 eV) can be identified as C≡N, the peaks at 284.7 and 288.8 eV as C-C and O=C-N, respectively. The N 1s peaks were also deconvoluted into three peaks at approximately 398.3, 400.5 and 402.6 eV. According to the N(E) CKVV spectra the surface samples contain an equal number of sp² and sp³ bonds of carbon atoms.

A study of electric field dependent spectra and the quantum electroluminescence yield of single-layer (SL) and double-layer (DL) light-emitting diodes (LEDs) based on a combination of an aromatic diamine (TPD) and an oxadiazole derivative (PBD) was carried out. It was shown that a high electric field weakens the recombination of Coulombically correlated electron-hole pairs in the bulk (SL LEDs) and enhances the process at the TPD-doped polymer/PBD interface (DL LEDs). The effects are discussed in terms of electric field mediated interplay between populations of localized (monomolecular) excitons, exciplexes and electroplexes in conjuction with their local environments.

Ordered arrays of cobalt nanowires have been fabricated by electrodepositing the corresponding materials into the pores of anodic aluminum oxide (AAO) membranes. Magnetic measurements illustrate their easy magnetization axis is along the nanowire. This kind of array can be used as a vertical storage medium. The ideal storage density is about 170 G bit in^-2.

We have conducted comprehensive measurements on enhanced electron attachment to ArF and KrF laser-excited benzene in the presence of Ar and N₂ buffer gases. At both these laser lines, two-photon absorption leads to excitation of benzene to energies above its ionization potential. Such excitations have been shown to lead to a population of long-lived, core-excited high-Rydberg states in addition to the ionization of the molecule. Present measurements on the dependence of negative ion yield on laser fluence, benzene pressure, and applied electric field verify that the observed negative ion formation is due to the attachment of the photoelectrons to the concomitantly produced high-Rydberg states. Using a rate equation analysis, the electron attachment rate constant for the core-excited Rydberg states was estimated to be of the order of 10^-4-10^-3 cm³ s^-1. Laser photoionization cross sections were also estimated, and the cross section at the KrF laser line is in agreement with a previous measurement.

In this paper, magnetic iron fibres of diameter down to 2 µm were produced by using the bundle-drawing method, and iron fibre-epoxy resin composites were prepared by mixing epoxy resin with the iron fibres. The electromagnetic parameters of the iron fibre-epoxy resin composites were measured using the transmission/reflection coaxial line method at microwave frequencies. The measured results show that the iron fibre-epoxy resin composites exhibit larger values of µ' and µ'' in the frequency range of 2-18 GHz, and the values of both ε' and ε'' steeply increase with increase in the fibre concentration. Based on a theoretical calculation of the reflection coefficient, the microwave absorbing properties of the iron fibre-epoxy resin composites were examined. The calculated results indicate that the iron fibre-epoxy resin composites exhibit good absorption performance in the radar band. The optimum absorption can be achieved by optimizing the fibre concentration. The absorption peak frequency can be manipulated easily by changing the thickness of the material layer. The calculated results also indicate that magnetic iron fibres may be useful in producing thin and lightweight radar absorbing materials.

The present paper applies both the first and the second laws of thermodynamics to the determination of the characteristics of an ideal diatomic gas heat engine that operates via a closed-loop circular cycle (viz of circular appearance in a P-V diagram when the pressure and volume units are appropriately scaled). The second law of thermodynamics reveals that an ideal gas undergoing an irreversible thermodynamic process P = P(V) with negative slope can reach a terminal point at d(PV^γ)/dV = 0 without ever going to equilibrium with its environment. Furthermore, heat flows into the gas when d(PV^γ)/dV>0 and out of the gas when d(PV^γ)/dV<0. These conditions are then used to determine where Q = Q_in and where Q = Q_out for a circular work cycle. For a specimen engine, the efficiency is then obtained from the first law of thermodynamics.

The gliding arc discharge (`Glidarc') is the subject of renewed interest in application to a variety of chemical reactions. The gliding arc creates a weakly ionized gas `string' between two horn-shaped electrodes. In this paper, we present a simple model for a bi-dimensional dc Glidarc working in air, in which the conducting zone of the discharge that is heated by the Joule effect is considered as a hot wire cooled by an air flow. Inside this wire, the heat transfer results from thermal conduction. The exchange of heat between the hot wire and the air flow is assured by convection and depends on the wire radius and the relative velocity of the arc with respect to the gas flow. The model correctly describes experimental results and allows us to predict the working parameters of the Glidarc in different experimental situations.

The influence of pulsation instabilities of the plasma torch on the radiation intensity and on the radial temperature profile was studied. Two limit cases of pulsations were considered: (1) the plasma is stationary in space, but has pulsations of temperature at each point of the torch and (2) the axis of the torch accomplishes space oscillations in the laboratory system of coordinates. For these cases the lateral and radial intensities and the radial distributions of temperature were calculated for a plasma with cylindrical symmetry. The calculations were made for the stationary and non-stationary model supposing that the pulsations follow a Gaussian distribution. For this study the spectral line of Ar I at 415.8 nm was chosen. The effect of the pulsations was studied for values of the different standard deviation. The results show that the pulsations of the plasma can significantly influence the experimental results and can be the cause of differences between the theoretical calculations and the experimental measures.

In nitrogen, for pressures below 200 Torr, the anodic glow is characterized by current oscillations, superimposed on a dc component. These oscillations may be attributed to fluctuations of the space charge, structured as a double sign sheath (double layer). The transition from a dark to a glow discharge can occur either directly from the oscillating state or, in the 20-200 Torr pressure range, through the appearance of a particular regime characterized by recurrent impulses. Both cases are considered here in the case of pure nitrogen, and the influence of pressure and gap length on the current and light waveforms is studied. An analysis of the light emitted by the discharge shows that a luminous structure, formed in the anode region, propagates towards the cathode, and travels a few millimetres before being absorbed by the sheath structure. This may be interpreted as the beginning of an ionizing front, which cannot propagate and is choked by the double layer. For higher applied voltages, the double layer cannot be maintained, and the current limiting effect seems to be suppressed during the time for charge evacuation; the current and light impulses corresponding to this latter case are then integrated into the much larger ones which characterize the glow discharge.

The first absolute density measurement of N₂(A ³Σ⁺_u;v = 0) metastable molecules in the short-lived nitrogen afterglow of a 440 Pa microwave discharge is achieved by using a intracavity laser absorption spectroscopy technique. The results obtained show that this technique is very well suited for the diagnostics of N₂(A ³Σ⁺_u) in the nitrogen afterglow. This is a first important experimental step with the aim of probing regions where the densities are expected to be the most significant. The densities measured in the discharge zone and in the short-lived afterglow maximum emission zone are about 5×10¹⁷ and 6×10¹⁶ molecule m^-3, respectively. The gas temperature is also deduced both from the Doppler profile of the spectrally resolved rotational lines and from the rotational distribution in the N₂(A ³Σ⁺_u;v = 0) state. These temperatures are about 1000 K in the discharge zone and 530 K in the short-lived afterglow maximum emission zone, respectively. The relatively high density of metastable molecules in the short-lived afterglow maximum emission zone reveals the important role played by these species in the short-lived nitrogen afterglow and can give a new insight into the mechanisms involved in the so-called nitrogen short-lived afterglow emission. We also conclude that this high density of N₂(A ³Σ⁺_u) in the short-lived afterglow results from an equilibrium between some local production mechanisms and destruction by reaction with N atoms.

A three-dimensional (3D) argon arc plasma at atmospheric pressure is presented. The model is developed using the commercial code Fluent. Two arc plasma configurations are studied: a free burning arc and a transferred arc. In the free burning arc configuration, the 3D results are compared with a two-dimensional (2D) configuration. The natural axis of symmetry of the system, given by the physical operating conditions and by the cathode tip geometry, leads to a good agreement between the 2D and 3D results. The model is validated by comparisons with experimental and theoretical temperature fields. The results show a small difference between the 2D and 3D models, as does the other literature. In the transferred arc configuration model the injection is given by three injectors. A rigorous modelling of the injection cannot be treated using a symmetric plane or a symmetric axis, so the results are presented in a real 3D configuration and compared with results obtained in a 2D configuration. Indeed, in a 2D configuration, due to the symmetrical condition on the axis, the main assumption consists of representation of the real injection geometry by an equivalent surface injection. The imposed mass flow rate is then distributed along a ring and differences can appear in the results.

An improved finite-element flux-corrected transport (FE-FCT) scheme, which was demonstrated in one dimension by the authors, is now extended to two dimensions and applied to gas discharge problems. The low-order positive ripple-free scheme, required to produce a FCT algorithm, is obtained by introducing diffusion to the high-order scheme (two-step Taylor-Galerkin). A self-adjusting variable diffusion coefficient is introduced, which reduces the high-order scheme to the equivalent of the upwind difference scheme, but without the complexities of an upwind scheme in a finite-element setting. Results are presented which show that the high-order scheme reduces to the equivalent of upwinding when the new diffusion coefficient is used. The proposed FCT scheme is shown to give similar results in comparison to a finite-difference time-split FCT code developed by Boris and Book. Finally, the new method is applied for the first time to a streamer propagation problem in its two-dimensional form.

Nitric oxides (NO and NO₂) and SO₂ emissions are a major environmental problem because of their negative influence on human health and vegetation. The federal regulations on limiting the pollution emitted by the engines of motor vehicles have triggered intense research on new techniques for the removal of these pollutants. New methods for exhaust gas cleaning are needed and among the several approaches to reduce the pollutant emissions, the non-thermal plasma technique shows promise (Luo J, Suib S L, Marques M, Hayashi Y and Matsumoto H 1998 J. Phys. Chem. A 102 7954).

In this work, a volume-averaged model is presented that can describe the removal of NO_x by the multi-pulse treatment of the exhaust gases at low temperatures and at atmospheric pressure in corona reactors. The model takes into account the production of active radicals after every discharge and the removal of NO by these radicals. Furthermore, the effect of ethene, one of the most important unburnt hydrocarbons in the exhaust gas, on the removal of NO is also investigated in this study. The effect of ethene has been investigated experimentally by several authors (Mizuno A, Chakrabati A and Okazaki K 1993 Non-Thermal Plasma Techniques for Pollution Control ed B M Penetrante and S E Schultheis (Berlin: Springer) p 165, Prather M J and Logan J A 1994 Proc. Combustion Institute25 1513), but there are almost no studies which try to explain, in detail, the chemical processes in such a system. The detailed reaction mechanism used in this study consists of 443 elementary reactions and 50 chemical species. The results of our numerical simulations show good agreement with the experimental results published in the literature. Reaction flow analysis and sensitivity analysis are performed in order to identify the specific reaction paths and the rate-limiting reactions for typical operating conditions of pulsed corona reactors.

A comprehensive self-consistent description of the column plasma of an inert gas dc glow discharge is presented. The description is based on the radially inhomogeneous kinetic treatment of electrons, ions and excited atoms in the plasma and includes the self-consistent determination of the radial space-charge field and the axial heating field. The determination of the axial field becomes possible by the additional consideration of the most important plasma-wall interaction processes and a balance of the electron surface density on the dielectric tube wall. The results of the description are compared with those of Langmuir probe measurements which makes conclusions about the properties of the plasma-wall interaction processes possible. It is shown that the wall processes and, hence, the wall properties are of great importance for the establishment of the electric field and the kinetic structure of the plasma components in the column.

It is shown that the constant current method, in which a dielectric is charged with a constant current while the voltage is monitored, allows one to determine the dependence of the stable ferroelectric polarization with the electric field. The determination is based on two successive experiments separated in time by a short-circuit period: a charging process in which polarization switching occurs followed by a recharging with the same current polarity. Analysis of the recharging experiments for poly(vinylidene fluoride), PVDF, shows that the polarization appearing in it is a metastable ferroelectric polarization, due to the reorientation of ferroelectric polarization lost during the short-circuit period. The method was applied to measure the ferroelectric polarization in PVDF samples with different β-phase contents and in an exploratory way for a few other ferroelectric polymers.

We evaluate, theoretically, the solar absorptance and thermal emittance of cermets with large particles. In particular, we consider Co and Ni particles with radii from 0.05 to 0.13 µm, embedded in binders of alumina and silica, respectively. For these particle sizes, it is necessary to consider the effect of the multiple scattering of light in the material for the visible and near infrared parts of the spectrum. The response of the particles is calculated using the Lorenz-Mie theory, and the multiple scattering effects are taken into account through a four-flux radiative transfer model. In the medium infrared, the Lorenz-Mie theory cannot be used because of the high absorption of the binder. The reflectance of the cermets in this regime is calculated using the Maxwell-Garnett effective medium approximation. In general, moderate values of the absorptance-to-emittance ratio are found.

A low-energy expansion of the EGS4 Monte Carlo code system was used in studies that aimed to understand photon production in transmission target x-ray tubes. In each case, foil targets were modelled with incident electrons of a few tens of keV. The production of bremsstrahlung and fluorescence radiation at different depths was studied and the angular distribution of radiation analysed. The characteristics of x-ray spectra measured along the beam path, for targets of different materials, thicknesses and several beam energies have been determined and the relation between efficiency and the fluorescence-to-bremsstrahlung ratio is shown.

We investigate the possibility of growing a uniform binary compound crystal in space, proposing a new crystal growth method. We develop a numerical calculation method for the growth of binary crystals, in which convection induced by temperature and concentration differences in the solution is taken into account. How to determine the shape and movement of the solution-crystal interface during the crystal growth is clearly explained for binary crystals, which is more complicated than for single-component crystals. The boundary fit method is employed to solve this moving boundary phase transition problem. The calculation method is applied to the crystal growth analysis of an InAs-GaAs binary semiconductor and the effect of buoyancy convection induced under microgravity conditions on the crystal growth process is investigated. It is found that the concentration field is disturbed and, as a result, the solution-crystal interface is deformed by buoyancy convection, even when the gravitational acceleration is as low as 10^-6 g, which is supposed to be the gravity level in the International Space Station which will start operation in 2004. It is also found that the direction of the residual gravity has a strong effect on the concentration field in the solution and the crystal growth process. Next, we analyse the influence of g-jitters and the Sorét effect on the crystal growth process. In fact, it is found that g-jitters and the Sorét effect have little influence on the macroscopic crystal growth process. The dependence of the generation of supercooling in the solution on convection is also investigated. It is found that supercooling is not induced by convection if residual gravity is 10^-6 g. Finally, we discuss the possibility of growing high-quality InGaAs crystals of uniform compositions in space.

In this paper some text and equations are incorrect. Corrected versions are given below.

In equation (20):

Also, the annotation for the horizonal axis in figure 1(d) is `Group velocity (km/s) At the angle of 30° to [100]'.