Table of contents

Extensions of the electrostatic probe technique to situations with negative ions, both theoretical and experimental, are described. Charged particle distributions and the plasma-sheath boundary are discussed in relation to probe measurements. Plasmas with negative ions, their occurrence and interest, particularly situations where the negative ion density becomes comparable to or exceeds the electron density are described. Some data in negative ion containing plasmas are shown and discussed.

A Poisson point process (PPP) presents a series of characteristics which may be used to define functions with an optimal signal-to-noise ratio. This work describes a statistical function, obtained by measuring time intervals in a PPP, which contains the same information about the signal to be analysed as the autocorrelation function. The advantage of this function is that it can be applied to cases of low light level in which the autocorrelation function gives a low SNR. As a particular case of low light level signals, the authors have used this method in a real experiment of single photon decay spectroscopy and measured the lifetime of the ⁴T₁ state of Mn²⁺ in an undoped TMMC crystal at room temperature.

Electron impact dissociative excitation of HgI₂ molecules producing the B² Sigma ⁺_1/2 state of HgI has been studied in the case of low energy electrons. Using the most intense band of the (B² Sigma ⁺_1/2-X² Sigma ⁺_1/2) system of the HgI radical at 444 nm, emission cross sections have been measured in the electron energy range 1-100 eV. The threshold electron energy for the observation of the (B² Sigma ⁺_1/2-X² Sigma ⁺_1/2) emission band system has been determined to be 5.7 eV. The role of such dissociative excitation in lasing was analysed.

The authors report the first study of reaction mechanisms on optically pumped vapour mixtures of HgX₂ (X=F, CI, Br and I) using ArF excimer laser excitation at 193 nm. Chemical and photolytically induced reactions take place yielding new species such as mercury interhalide complexes (HgXX'), interhalogens (XX'), dimers (Hg₂X₂), 'heavy' mercury (II) interhalogen complex dimers (Hg₂XX'), as well as free mercury (Hg) and halogens (X₂). The effects of such mixtures upon the laser absorption and emission bands for both photolytically pumped and electronically excited laser emission and the production of interhalogen compounds and their effect on laser emission is discussed.

The authors show that the spectroscopic monitoring of a pulsed discharge plasma is a powerful tool for investigation of the kinetics of elementary processes. Such studies allow discrimination between different collisional processes which cannot be distinguished in cw operations. The experiment is applied to the helium-cadmium discharge successively created in positive column and hollow cathode devices. The main processes responsible for the laser emissions are pointed out. Comparing the results obtained in two different mixtures: Ne-Cd and He-Cd, it is shown that the Cd⁺(6G) level emitting the 636.0 nm and 635.5 nm laser lines is mainly populated by He⁺/Cd charge transfer reaction. The green 534 nm and 538 nm laser lines are partly due to the 6G radiative decay and partly to the Cd⁺ ground state excitation by electron collisions. This last process is also responsible for the Cd⁺(5s²²D₅2/) level emitting the 441.6 nm laser line. The Penning ionisation process is another population channel of this level. The results concerning the blue and red line emission sources are now well established whereas those related to the green one are more original. Furthermore, the kinetics of different electron energy ranges are studied in time. This means that information on the electron energy distribution function can be spectroscopically obtained during the pulsed excitation.

In order to study the mixing between a cooling plasma jet and a cold gas flow, an axisymmetric laboratory scaled configuration is investigated. The mixing is followed by recording the averaged temperatures. Data processing allows the establishment of an empirical law linking the mixing length and the parameters of the system. Through the application of the turbulent numerical model HADES elaborated by the Laboratoire National d'Hydraulique at Chatou (France) to the studied configuration, some useful conclusions for plasma flow modelling are proposed.

Electron energy distribution functions (EEDF) in Ne-Xe-HCI mixture have been calculated by solving the time dependent Boltzmann equation in the presence of a sinusoidal field E(t)/p₀=(E₀/p₀) cos omega t. The results have been obtained for different reduced field frequencies omega ( pi *10⁴<or= omega <or= pi *10⁷ s^-1 Torr^-1), while the field amplitude E₀/p₀ has been self-consistently calculated by imposing an electron balance during the period between the gain by ionisation processes and the losses by dissociative attachment and ambipolar diffusion. Different fixed concentrations of excited states have also been considered to understand the role of superelastic collisions as well as ionisation and attachment from excited states in affecting EEDF. The results at the fixed omega /p₀ show that the presence of excited states strongly decreases the modulation of EEDF enlarging its tail at the same time. Moreover the ionisation frequency from excited Xe* is such as to decrease the self-consistent field amplitude necessary to sustain the electrical discharge. Finally the results for different field frequencies show that a complete quasi-stationary situation is only reached at Gw approximately pi *10⁴ s^-1 Torr^-1.

Room temperature measurements of the ratio D_T/ mu between the transverse diffusion coefficient and the mobility of mass-identified ions in a neutral gas are reported for Na⁺ ions in argon and neon. A new apparatus was used which, in contrast to previous methods, does not involve mechanical movement which alters the geometry of the drift tube assembly during a measurement scan, or variation of the experimental conditions inside it. Results with an estimated accuracy of better than 3% were obtained at E/N values ranging from 10 to 260 Td with argon and 10 to 200 Td with neon.

The need for more basic understanding of welding processes has engendered efforts to provide experimental data for use in weld process modelling. To this end, the spectral analysis of a welding arc (TIG) is described. The effects of the anode spot on metallic emission are presented. Measurements have been performed beginning at 0.5 mm above the workpiece and show that the welding pool has no effect on arc properties and that the LTE hypothesis is valid near the plasma axis.

The application of the soft X-ray backlighting technique to measure the density of a dense plasma produced in a capillary discharge is demonstrated. A laser produced tellurium plasma was used as the X-ray source and a two-dimensional flash image with a 140 mu m resolution was obtained. From the X-ray absorption along lines of sight and using the Abel transform, assuming a cylindrical symmetry, the radial density profile was deduced. The maximum plasma density measured was 6*10¹⁹ ions/cm³.

A numerical simulation including transport phenomena is proposed for the calculation of ozone production in a positive point to plane corona discharge. The influence of pressure, temperature and average flow velocity have been investigated. The results show that the average flow velocity must be adjusted to obtain the maximum ozone concentration (optimisation of the kinetic processes). The respective influence of kinetics and diffusion processes is also analysed in the limits of low pressure (5 Torr) and high pressure (760 Torr).

The anisotropic thermal expansion of LiB₃O₅ has been studied along the principal crystallographic directions over the temperature range of 17 to 790 degrees C by means of high-temperature X-ray powder diffraction. This orthorhombic crystal exhibits strongly anisotropic expansion with coefficients of 108.2*10^-6 K^-1 for (100), 33.6*10^-6 K^-1 for (010) and -88.0*10^-6 K^-1 for (001) at 50 degrees C. The coefficients tend to zero when the temperature approaches the crystal's peritectic reaction point.

The behaviour of oxygen impurities during rapid thermal processing of Ti/Si diffusion couples has been studied between 480 and 800 degrees C. Samples were prepared by sputter deposition of 40 nm of titanium of (001) silicon wafers which were then annealed in ambient argon. The investigation techniques included depth profiling using Auger electron spectroscopy, transmission electron microscopy, and electrical sheet resistance measurements. The oxygen impurity effects are mainly controlled by kinetic parameters, i.e. the silicide reaction rate and the oxygen diffusivity in titanium. At low temperatures, the oxygen contamination originating from the annealing furnace is the cause of major impurity effects. Below 550 degrees C the oxygen that spreads into the metal film blocks the silicide reaction while above 650 degrees C, silicide formation dominates over oxygen diffusion and contamination-free silicide is produced. From 550 to 600 degrees C, the competition between silicide reaction and oxygen diffusion results in the formation of a silicon-deficient, oxygen-rich Ti-Si-O sublayer on top of the silicide. This layer appears to be composed of TiO_x and TiSi₂. In some cases, all Ti is reacted but oxygen remains in the layer and the morphology of the silicide is strongly affected. In any case, the only crystalline silicide phase that forms is TiSi₂, and not TiSi or Ti₅Si₃, as confirmed by electron diffraction and high-resolution electron microscopy.

Charge transfer from metals to silica (and SiO₂ grown on silicon) is predominantly negative and significantly dependent on the workfunction of the contacting metal. Charge transfer to soda glass, on the other hand, is strongly positive and does not vary much with metal workfunction; these features are unchanged by a wide range of surface treatments and are clearly intrinsic to the material itself, not an accidental result of some particular surface condition. The authors present strong evidence indicating that the positive charging of glass is a consequence of the presence of sodium (that is, in ordinary glass; other metals such as silver can be similarly effective). One consequence of this is that the charging of glass can be strongly modified by exposure to water, which leaches out the metallic ions. The presence of mobile sodium ions also causes pronounced charge accumulation when contacts are repeated to the same spot, an effect which is virtually absent from silica. Their observations suggest that the amount of charge transfer to glass must be limited by some factor other than the density of electron donor sites, e.g. by transfer of electrons or ions during separation.

In order to study the effect on the frictional electrification of molecular orientation in surfaces, the polarity and magnitude of the charge generated on thin films of stearic (SA) and oleic (OA) acids deposited as a Y-type assembly by the Langmuir-Blodgett (LB) method and on polyethylene (PE) and polymethyl methacrylate (PMMA) films were investigated when the films were rubbed by a metal ball. The polarity of charge, for example, on the three-layer LB film of SA was the same as that on PE film. The electrification charge on LB films, in which the orientation of molecules is thought to be controlled perfectly and the molecular density is very large, was larger than that on PE films, on the surfaces of which many defects or disorders in molecular orientation exist. The electrification results were compared with the normal load dependence of frictional force measured simultaneously. It was concluded that the penetration of the metal ball into the molecular aggregate, the regularity of molecular orientation and the number of molecules per unit area determine the polarity and magnitude of frictional electrification charges.

Glass-metal nanocomposites involving iron, nickel and copper in a silica glass matrix have been prepared as films on glass substrates by the sol-gel route. The metal phase is introduced as chloride and silicon tetraethoxide is used as the glass precursor. DC electrical resistivity of these nanocomposites has been measured over the temperature range 100 to 350 K. Resistivities varying from 10 Omega m to 10⁶ Omega m have been obtained by controlling the volume fraction of the different metal phases within the silica glass. Low-temperature resistivities of the samples are controlled by a simple activation with energies in the range 0.01-0.1 eV. At temperatures above 150 K some of the nanocomposites exhibit another activated mechanism. The latter is believed to arise due to hopping of electrons between the localised states formed by the distributed metal atoms within the silica glass matrix.

Stimulated Raman spectroscopy can be used to monitor the concentration of species in an injected liquid fuel spray. A method based on the relative intensities of two adjacent Raman peaks is developed for determining concentrations in an evaporating two-component fuel. Measurements in evaporating fuels containing two different boiling fractions are reported.

Recent experiments in stimulated Raman scattering in droplets suggest that the enhancement of the nonlinear gain can be phenomenologically modelled by considering the pumped active path to be greatly enhanced. Total internal reflection traps the signal in the droplet at morphology-dependent resonances. A simulation of this and a number of other new aspects of the process that result from considering a variable pumped gain within the droplet are described.

Solar control coatings used in architectural glazings of buildings in warm climates should exhibit controlled optical transmittance (10-50%) in the visible region and must reflect effectively (90%) in the IR region. Physical vapour deposited CdSe_xTe_1-x thin films were studied for solar control applications. The films were characterised with respect to film parameters such as film thickness, substrate temperature and deposition rate to get films with good solar control characteristics. The solar control characteristics of the films were calculated in the range 0.40-2.40 mu m. It was observed that the solar control parameters of the films are composition dependent but depend also on the above film parameters. CdSe_xTe_1-x thin films of thickness about 0.1 mu m exhibited better solar control characteristics than metallic coatings and other films such as Cu_xS and PbS.

The early stage of clustering and the subsequent precipitation of Si atoms in Al+1.37, Al+2.83 and Al+6.81 at.% Si alloys have been investigated by means of electrical resistivity and microhardness measurements. The results showed that: (i) the interaction between the vacancy-type dislocations and the partially precipitated Si atoms is the predominant process during the initial stage (25-150 degrees C); (ii) as the large particles of Si precipitates are formed, they become no longer coherent with the Al matrix; (iii) the microhardness of the as-quenched specimens was found to increase linearly with increasing Si concentration in the alloy.

Some reasons are discussed for the stabilisation of a glow discharge at atmospheric pressure which was attained by controlling the following three conditions: the use of a high-frequency source, the use of He gas for dilution and the insertion of a dielectric plate between electrodes. The three conditions interact: the dielectric plate in a plasma forms the pulsed discharge from a low-frequency source, the fast duration of pulse current prevents a transition to an arc style discharge and a large volume of metastable atomic helium aids ionisation or dissociation near the electrode plate and in the flowing gas. The measurements of discharge-maintaining voltage at several hundred volts provide evidence that this is really a glow plasma at atmospheric pressure. The changes of the emission intensities of metastable helium prove indirectly that some dissociation occurs as a result of the action of the helium atom.

A method for non-contact ignition of DC arcs to eliminate RFI and maintain safety by using a continuous sinusoidal HF voltage has been investigated. Characteristics of combined AC-DC discharges showed that, during ignition, three types of discharge exist: AC dominated, DC dominated and transition region. To achieve ignition of the arc, the DC-dominated discharge must be reached, which is possible when the main DC supply load line cuts only the DC-dominated curve. For the ignition of a 3 mm (tungsten inert gas) TIG arc supplied by a typical DC welding supply of 80 V output, the HF ignition source should have an output voltage of at least 2.5 to 3 kV and output current of about 0.9A.

Experimental data on the isothermal decay associated with peak 4 of LiF:Mg,Cu,P are presented. The decay has been measured at various temperatures in the range 170-190 degrees C, presenting always the purely exponential variation of the light intensity with time typical of a first-order kinetics process. From the good linear dependence of the logarithm of the decay lifetime (In tau ) with the inverse of the temperature (1/T), also typical of a first-order process, values for the activation energy and the frequency factor have been obtained for this peak. The isothermal decay features are discussed with regard to the results of Srivastava and Supe (1988) who claimed that peak 4 originated in two different groups of traps.