Table of contents

The method of maximum likelihood has been used to calculate the estimators, and their standard errors, of the parameters of joined half-gaussian fits to back-scattering data. The range and straggle of 60-400 keV As implants into Si at various doses have been measured with a precision of two or three nm and a comparable accuracy, with normal incidence RBS.

Methods of performing time domain dielectric spectroscopy have been evaluated using both exponential and a t^-n power law response. The exponential decay gave accurate results using both a complete transform and a numerical analysis algorithm based on overlapping parabolas over a time period truncated by typical experimental limitations. When this algorithm was used for the t^-n response, it gave rise to a constant error term in the real part of the susceptibility, xi '( omega ), which was independent of frequency. It thus appears that the xi '( omega ) spectrum cannot be obtained reliably using time domain methods.

Changes of physical and chemical properties of paint films resulting from accelerated weathering have been detected using a new measurement technique: opto-thermal transient emission radiometry (OTTER). Opto-thermal decay times were found to decrease with weathering exposure, the effect being most pronounced in the early stages of degradation. The technique promises to provide a useful means of monitoring paint film deterioration.

Some results are reported of static and impulse corona onset and breakdown voltage measurements in 10 mm point-plane gap in SF₆ mixture containing 0.5% and 1% by volume of Freon 113. It is found that the corona inception voltage in the mixture is higher than for pure SF₆ over the pressure range investigated (up to 4 bar) and that the corona stabilisation peak is greatly extended. Measurements of the minimum impulse breakdown voltage have shown that the improved stabilisation is associated with an increase in the leader propagation voltage in such mixtures. The results are of significance in connection with particle-triggered breakdown in commercial compressed-gas-insulated cables and sub-station equipment.

A photoacoustic (PA) spectroscopy measurement is performed for F₂ centres in X-irradiated LiF single crystals. The observed half-width of the PA spectrum is noticeably larger than that of the optical absorption spectrum, although the optical density is not so high as to induce the saturation of the PA amplitude expected from the 'thermal piston model' of Rosencwaig and Gersho (1976) which is usually used. A possible explanation is made for the half-width broadening based on the 'composite piston model' given by McDonald and Wetsel (1978).

The influence of doping a AgBr emulsion with OsCl₆^2- on the transient photoconductivity at 77K is investigated. The addition of OsCl₆^2- as a dopant in AgBr emulsion grains introduces an electron trap and, simultaneously, enhances recombination.

The first measurements of conductivity in the temperature range 25-300K for conductive systems made by crystallisation of charge-transfer complexes in polymer matrices are reported and discussed in terms of the metal-insulator transition in organic metals.

In SrTiO₃ (strontium titanate) and TiO₂ (rutile), the temperature dependences of the optical dispersion parameters (the dispersion energy E_d and the single-oscillator energy E₀ developed by Wemple and DiDomenico, 1971) were investigated in the 450-1600 nm wavelength range and at temperatures from room temperature to 400 degrees C. E_d and E₀ exhibit no temperature dependence in SrTiO₃. The temperature dependence of E_d in TiO₂ (both ordinary and extraordinary cases) are negative, and E₀ exhibits no temperature dependence in TiO₂.

The temperature dependence of the linear electro-optic coefficient r= mod r₁₃-r₃₃ mod of LiKSO₄ at a wavelength of 6328 AA was measured in the temperature T range of T>-62 degrees C (phase I) and -62 degrees C>T>-86 degrees C (phase II). The value of r is 1.9+or-15*10^-12m V^-1 at 20 degrees C and is almost constant in phase I, and the rapid increase in r with decreasing temperature near the phase transition temperature of -86 degrees C in phase II was observed.

Changes in the infrared spectra of polymers are known to occur when an electric field is applied. To determine if this is due to mechanical deformation, measurements were made on low density polyethylene film over a range of applied stress. Below the plastic limit the only effects seen were attributable to thickness changes. It is reasoned that changes in the infrared spectrum in an applied field are most probably associated with space charge formation in the polyethylene.

A method of producing thin films of ITO is presented. DC sputtering deposition was done in the following way: the In_0.88Sn_0.12 target was oxidised by glow discharge in a pure oxygen atmosphere (15 Pa); after the removal of oxygen from the deposition chamber, the sputtering process was done in an argon-oxygen atmosphere (5.3 Pa). The substrate temperature was held at 453K. The authors have obtained 450 AA thick films with resistivity of 6.5*10^-4 Omega cm, Hall mobility of 12 cm² V^-1 s^-1 and achromatic characteristics in the visible region. The films had reproducible and stable parameters. Such films can be used successfully as transparent electrodes for liquid crystal displays.

Equations for the reflection coefficients associated with radiation scattered by dispersive media are obtained. The Pade approximant technique is used to calculate the reflection coefficients. The results for the (0/1) Pade approximant lead to computationally useful results that compare well with the exact results.

The electromagnetic field scattered by a perfectly conducting elliptic cone is discussed. The problem is three-dimensional and is presented by the dyadic Green function, expressed by the help of Lame periodic and nonperiodic functions. Numerical results for the field and the current distribution are presented for some special cases.

A theoretical interpretation of the digital beam deflection efficiency of an electro-optic modulator is described. Calculated switching voltages are in good agreement with the experimentally observed values. The computed percentage efficiencies at three successive positions are 57, 48 and 43.

Expressions are presented for the determination of the activation energies, E, of first-order peaks of thermoluminescence of thermally stimulated depolarisation currents, using the widths or half-widths of the peaks. These widths correspond to temperatures at which the signal level is ¹/₄, ¹/₂ or ³/₄ of the peak height, on either side of the peak temperature, T_M. Very exact expressions are derived, valid in the region of E/kT_M values between 10 and 100. Also given are expressions for the widths of the peaks when their energies and peak temperatures are known.

It is shown that by using an optical fibre of 'tubular' structure, comprising a thin layer of higher refractive index sandwiched between a core and a cladding of lower refractive index, it is possible to transmit the TE₀₁ mode in isolation with the HE₁₁ mode virtually cut off. Possible advantages of this arrangement are discussed.

The analytical shape of the ion distribution function of a weakly ionised, weakly dense gas subjected to the action of an intense electric field is discussed. The authors show that the Grad expansion of the solution of the Boltzmann equation converges slowly and that a very high order of expansion is required. A generalised polynomial solution is presented and applied to the solution of Boltzmann's equation. The variation of the shape of the stationary distribution function with the electric field strength is discussed.

A general expression is given for the ambipolar diffusion coefficients in a gas discharge plasma containing an arbitrary mixture of electrons and positive and negative ions. Specific examples of coefficients for three-component active discharges and afterglows (electrons, one type of singly charged negative ion, one type of singly charged positive ion) are presented. The general expression is also used to clarify the limitations associated with the coefficients for a three-component plasma published by Thompson (1959). The limitations are significant.

With the exception of simple Langmuir probes, all other probes used to diagnose the scrape-off plasma behind the limiter radius of separatrix in Tokamaks are usually sufficiently large that they disturb the plasma they are measuring. An analysis procedure is outlined for recovering the undisturbed plasma parameters from those measured by a large probe in a collisionless scrape-off layer. The analysis is applied to experimental data recorded by large probes inserted into the scrape-off layer of the DITE Tokamak.

A method for computing the radial temperature distributions in wall-stabilised arcs is developed and is applied to the treatment of D-line emission from a sodium arc lamp. The method adopts a radial temperature distribution containing two free parameters whose values are derived from two coupled nonlinear equations, one describing the energy balance on the arc axis, and the other the integrated energy balance. The radiative terms in the energy balance equations are obtained from a solution of the exact transfer equation.

The absolute infrared calibration of the Si_Ga centre in GaAs has been determined using samples analysed by atomic absorption spectrometry. It is shown that an absorption strength of alpha Delta =1 cm^-2 due to the band at 384 cm^-1 wavenumbers corresponds to a Si_Ga concentration of 6.0+or-1.0*10¹⁶ cm^-3. Using grating IR spectrophotometers the measurement of this band in highly doped samples is found to be underestimated when the transmission falls below about 10%. It is shown that this photometric error arises from the measurement of a strong sharp line with instruments of limited resolution. A residual two-phonon absorption feature in this spectral region is present in undoped GaAs and this restricts the lowest Si_Ga measurable concentration to 5*10¹⁵ cm^-3 when dispersive instruments are used.

A virtually complete picture of the substitutional-interstitial diffusion mechanism occurring in semiconductors is described. The mechanism is modelled on the computer using a numerical technique without attempting to employ differential equations. The interstitial impurity atoms are taken to diffuse with a finite diffusion coefficient. Interstitial diffusion and self-diffusion by atoms of the host crystal are taken to occur independently and do not interfere with each other. The system is allowed to approach towards equilibrium after interstitial and self-diffusion within an iteration time. The effect on the diffusion profiles of the variation of such parameters as diffusion time, diffusion coefficients and the bulk generation of vacancies are studied. Results obtained from this work are consistent with published experimental diffusion work on semiconductor systems.

The concentration distribution obtained from a two-step diffusion process in glass under DC electric fields has been theoretically estimated. The theoretical profiles have been favourably compared with experimental results from the literature.

The present study has examined the relationship between the I-V electrical characteristics and the interfacial microstructure present in Ag/In-(100)InP metal-semiconductor contacts. By depositing mixed alloy contacts of silver and indium on to (100)InP surfaces and then using thermal treatments to activate the interface, a variety of different microstructures has been generated within the interface of the contact. The different microstructures have then been related to the electrical characteristics of the contact. It has proved possible to engineer the contact's interface into differing states that reproducibly generate Schottky or truly Ohmic-like behaviour. The alloy contacts have been deposited with precise compositions by making use of separate Knudsden sources operating within an MBE system. The contact's interface has been characterised with Auger profiling and by direct observation with scanning electron microscopy.

Surfaces of 10 Omega cm, n-type (111) silicon wafers were oxidised in trychloroethylene (TCE) ambient. MOS test devices were prepared to characterise the Si-SiO₂ interface by AC field-effect and C-V methods. The AC field-effect mobility mu _FE as a function of frequency showed a small temperature dependence of relaxation times (8-4 mu s) in the temperature range 240-370K. The dominant energy level E_T was found to be 0.04 eV below the conduction band edge and the effective capture cross-section of interface traps was found to be about 10^-1 cm². C-V measurements on the same structure showed that the normal distribution of interface state density N_SS increased towards the conduction band edge from the mid-gap energy. The effect of TCE tends to reduce the density of interface states N_SS as well as shift the energy of shallow traps E_T towards the conduction band edge.

The plasma polymerisation of phenylacetylene (PA) was carried out at 13.56 MHz and its electrical properties and structure were investigated. The plasma-polymerised PA films were found to have chemical structure similar to that of chemically polymerised polyphenylacetylene in a short range of molecular chains. These films were found to be electrically insulating. They also showed substantial photoconductivity at long wavelengths such as 700 nm; this was found to arise from the excitation of trapped charges. The doping of the plasma-polymerised PA films with iodine increased both the photoconductivity and the dark conductivity.

A universal and relatively simple method is proposed for numerical recognition of the time correlation function of a system of interacting particles if its Mossbauer (absorption) spectrum is available. Applicability of the method is demonstrated for the case of Brownian motion of particles suspended in a fluid, by numerical simulation of the spectra. The effect of the statistical errors of measurements on the solution is also properly accounted for. Some results of previous authors are discussed.

The authors present details of the effect of modulation frequency variation upon the waveform of the detected photo-acoustic signal from ferromagnetic resonance in a thin film by both chopping and sinusoidally varying the microwave power level. The deviations from the predicted omega ^-1 dependence are explained for sine-wave modulation by considering losses to the sample environment. For square-wave modulation, distortion of the waveform and the loss of higher harmonics must also be included. The factors determining the relevance of the application of the photo-acoustic effect to the detection of ferromagnetic resonance are also reviewed.

The method of initial rise, as used for the determination of activation energies in thermoluminescence and thermally stimulated depolarisation currents, may result in energy values which are significantly lower than the correct ones if the levels of signal used are higher than a few per cent of the peak height. An expression is derived which gives the corrections to be added to the energy values found in this way.

In LiF:Mg, Ti dosimeters, gamma -irradiation at low temperature (77K) reduces the TL intensities of peaks 7 (260 degrees C) and 10 (450 degrees C) drastically whereas the TL intensity of peak 5 (200 degrees C) remains nearly the same as that after irradiation at room temperature (RT). This implies that during irradiation, interstitial migration is involved in the formation of traps giving rise to peaks 7 and 10. The generic nature of peaks 7 (Z₂) and 10 (Z₃) traps and their (Z₂ to or from Z₃) trap conversion are further demonstrated. Since diffusion processes are inhibited, F and Z centres are not formed efficiently in LiF dosimeters irradiated at 77K. Studies on supralinearity, sensitisation, thermal history, optical absorption, optical bleaching, phototransfer and LET dependence following irradiation at RT and at 77K imply that four different processes (competing nonluminescent centre, track interaction during read-out, Z centre and mobile interstitial) are operating simultaneously to varying extents during the trapping/luminescent stage of different TL glow peaks. The photon emission mechanisms of these peaks are discussed. The claim that supralinearity of peak 5 is related to Z₃ production is disproved. Further studies on Z centre identification in LiF:Mg, Ti are indicated.

The nitridation of the surface of titanium and zirconium samples was brought about by a powerful, microsecond pulsed TEA CO₂ laser irradiation in an atmosphere of technical nitrogen. A breakdown plasma was ignited and developed under the action of every incident laser pulse in the vicinity of the surface of the irradiated samples. There is evidence of strong competition between nitridation and oxidation brought about by the action of the breakdown plasma.

Excited atomic and molecular states of Ar, N₂ and H₂ have been detected in plasmas pulverising Ti for TiN deposition on steel substrates. Emission spectroscopy has been used to measure relative densities of species between the emitter and the substrate. Such variations can be correlated to the coating composition which continuously varies from Ti₂N close to the Ti emitter farther away.