Surface morphology, structural and optical properties of CuInS2 thin films grown by the single source thermal evaporation in vacuum have been studied. The films were annealed from 250 to 550 °C in argon atmosphere with low oxygen concentration (O2 < 2 ppm). CuInS2 films are characterized by scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDX), X-ray diffraction (XRD) and optical transmittance and reflectance spectra. The maximum grain size of the sample after annealing at 550 °C was over 150 nm. The electron dispersion spectroscopy and X-ray analysis concludes that the polycrystalline CuInS2 thin film after annealing above 200 °C were sulphur-poor. We obtain CuInS2 layers with high structural and optical quality at annealing temperature above 450 °C with formation of In2O3 as minority phase. The band gap energy of the CuInS2 films after annealing above 450 °C was about 1.50 eV which perfectly matches the solar spectrum for energy conversion.

High power RF device performance decreases as operation temperature increases (e.g. decreasing electron mobility affects cut-off frequencies and degrades device reliability). Therefore determination of device temperature is a key issue for device topology optimisation. In this work the temperature variation of AlGaN/GaN high-electron-mobility transistors grown either on silicon or sapphire substrate under bias operation was measured by micro Raman scattering spectroscopy. Temperature measurements up to power dissipation of 16 W for 4 mm development devices were carried out and a peak temperature of 650 K was determined. The difference of thermal resistance for similar devices grown on the two different substrates was assessed. The thermal resistances of different device topologies were compared to optimise the component design.

A new and simple process for the selective emitter realization was developed on multicrystalline silicon wafers. This material is in competition with single-crystal silicon since it is able to lead to similar performances with a reduction in the cost of solar cell realization.This work is centred on the study of emitter area of a photovoltaic cell and the possibilities to obtain a selective emitter in only one step while avoiding the use of chemicals.This would make substantial economies on the rejections treatment which became a capital environmental factor. A structure with selective emitter consists of a heavy doping under the metallic contacts, leaving weak the surface concentration between the grid lines.This allows a good surface passivation while keeping a good contact resistance for screen printed lines.The advantages of such a structure could be observed by quantum efficiency measurements yield where the benefit appears in the UV-VIS range of the solar spectrum.

The mechanism of electrical degradation and breakdown of polymericinsulating films is examined by this paper as a two-stage process.Electron injection by Fowler-Nordheim tunnelling, provides theenergy for the creation of the first cavity near to the injectioncontact (the beginning of the formation of the so-calledlow-density region) where impact ionisation of molecules and henceelectron avalanche, can occur. This stage is assisted by intrinsictunnelling of electrons through local potential barriers. Thesecond stage, the macroion bond scission and the creation ofanother macroion and a free radical is a thermofluctuationalprocess which involves the action of the stretching force by thelocal electric field. From the other hand breakdown initiationstarts when an electron following a Poole-Frenkel hoppingmechanism, is accelerated in a hole with sufficiently largedimensions. The role of the applied electric field F a and itsrelation to the local electric field F l in both stages isexamined. Experiments were executed for measuring the breakdown ACvoltage V b and dielectric strength F b , with voltagerising rate 3 KV/s, in order to examine their relation withspecimen thickness d and hence to derive a steady statebreakdown criterion. Measurements of PET films life-time were alsotaken in order to examine the degradation mechanism and thepolymer's lifetime.

This paper presents and discusses some numerical approaches for the simulation of induction machine behaviour under no-load conditions. All the approaches are based on the finite element solution of the electromagnetic field problem, including physical models of magnetic material and interaction with external circuits. The computed flux distributions enable the prediction of magnetic and additional losses, whose accuracy is discussed also in the light of the complexity of the used numerical technique.

For red, green and blue OLEDs that emit for each of them in a wide spectral range ( $\Delta \lambda \approx200$  nm), we calculate the luminous luminance, the energetic luminance and the various yields in the pseudo-monochromatic hypothesis (ideal source generally considered), and we compare with a calculation in relation with a polychromatic system (real source). Finally, we demonstrate that the values of the energetic luminance and of the energetic or quantum yield are practically unchanged whatever the hypothesis; on the contrary, in the simplified case of the pseudo-monochromatic source, the luminous luminance and the luminous yield of a green organic diode are overvalued in relation to the polychromatic real case, whereas for a blue or a red organic diode these parameters (calculated in the pseudo-monochromatic hypothesis) are undervalued. The error is all the more great since the OLED emission is far away from the maximum value of the photoptical eye response.

In this paper, the Finite Integration Technique and an approach to solve simultaneously the magnetic and electric circuit equations is presented in magnetostatic case. The developments of the FIT will be compared with those used for the Finite Element Method. As example of application an iron core coil is studied and the results of both approaches are compared.

Solid bodies obtained by sintering glass microspheres through the first stage of sintering have the useful characteristics of large porosity combined with adequate mechanical strength. Additionally, if the microspheres all have the same diameter, the pore morphology is well defined in terms of shape and has a narrow distribution in size parameters. The two major areas of application of the porous bodies are: A. light weight but mechanically strong objects, and B. filtration of liquids. The fracture strength and the specific liquid permeability of these bodies are expressed as empirical functions of volume porosity (ratio of empty volume to the total volume). Here, we describe the results of our systematic study of the geometrical properties like density, linear contraction and global porosity parameter φ, and the flow parameter, the permeability coefficient as a function of the sintering conditions. Our analysis shows that the existing relations, property versusφ,that claim to describe the behaviour are not adequate for our simple system. We think that the contact area between the adjoining spheres is a better parameter to consistently describe a larger set of the properties. Since, for the initial stage sintering of single diameter spheres, the contact area and the pore size can be expressed as a function of linear contraction, latter is a better universal parameter in terms of which the regular features of the property variations can be expressed.

A high-performance InP/InGaAs $\delta $ -doped pnp heterojunction bipolar transistor (HBT) has been first fabricated and demonstrated. The addition of a $\delta $ -doped sheet between two undoped spacer layers effectively eliminates the potential spike at emitter-base junction, lowers the emitter-collector offset voltage, and increases the barrier for electrons, simultaneously. Experimentally, a maximum current gain of 50 and a low offset voltage of 70 mV are obtained, respectively. To our knowledge, the offset voltage of the studied device is the lowest value for the reported InP/InGaAs pnp HBTs.

1.3 $\mu $ m room temperature emitting multiple-stacked InAs/GaAs(001) quantum dots (QDs) are grown by molecular beam epitaxy (MBE) and investigated by photoluminescence (PL), polarized photoluminescence (PPL), photoluminescence excitation (PLE), time resolved photoluminescence (TRPL) and atomic force microscopy AFM. The PL measurement shows that two distinct sets of QDs coexist in the sample. The AFM image of the tenth QDs layer not only confirms the bimodal size distribution of the QDs but also shows that the large QDs are elongated along the [1-10] direction. The former structural information has been verified by PPL. Through the excitation density dependent PL and the detection energy dependent PLE we have evidenced two kinds of QDs within the small size dots population: isolated QDs and laterally coupled QDs with vertically coupled large size QDs. The large size dot population is found to possess a long PL decay time confirming that they are electronically coupled. In the meanwhile the PL decay time of the small QDs is found to be similar to that of single layer QDs. These results would help improve understanding some fundamental properties of an interesting structure for optoelectronic applications.

We have investigated the influence of the steric properties of conducting particles in a nonconducting host matrix on the conductivity threshold of the material, i.e., the minimum volume fraction of conducting phase for the whole sample to become conducting. A statistical, numerical method is used in which the particles are randomly put, one by one, into the nonconducting host and the conducting path is searched. The particles are allowed to penetrate each other to some extent. Three different types of particle shapes are considered: spherical, cylindrical with rounded ends and asymmetric cuboids with rounded surfaces. We have found that in addition to the anisotropy in the particles' dimensions, the angular distribution of the particles' long axes plays a dominant role in the calculations of the conductivity percolation threshold.

The substitution sites of Pb and Dy dopants in the cuprate-type high temperature superconductor Bi2Sr2Ca1Cu2O $_{8+\delta}$ are determined by a direct comparison of the angle-scanned X-ray photoelectron diffraction (XPD) patterns. We demonstrate the power of XPD as a fingerprinting tool and conclude that Pb occupies Bi sites and Dy the Ca sites. The presence of the incommensurate lattice modulation is not visible in XPD, probably due to a Pb-induced, reduced modulation amplitude.

The different types of errors affecting the spectral purity of a rapidly chirped continuous wave laser are experimentally studied using an unbalanced interferometer. The response of the interferometer to the deterministic (periodic and aperiodic) and stochastic quantum and technical laser frequency noises is theoretically analysed and experimentally investigated with an external cavity diode laser operating at 793 nm and providing chirps in excess of 10 GHz on the ms timescale. The system is shown to be able to measure the laser frequency to better than 1 MHz during the chirp. It provides an error signal that can be used to servo-control the laser frequency in real time while it is chirped.

Pulsed laser ablation of metal samples in liquid environments by using a combination of two laser beams has been examined in order to clarify a possibility of the formation of size-selected nanoparticles with controlled dimensions. It has been shown that double-pulse laser ablation technique in transparent liquids can be a good choice for fabrication of stable metal nanoparticles where an efficiency of ablation process can be increased and the wider range of a controllable variation of the mean diameter of the fabricated particles can be achieved.

Fe $_{1-x}$ Crx thin films on Si and fused-SiO2 substrates were prepared using ion-beam sputtering deposition. Films with x = 0.69 and x = 0.49 were characterised by X-ray diffraction and conversion electron Mössbauer spectroscopy experiments, and concurrently the elastic and plastic behaviours were investigated through indentation tests. For x = 0.69, the paramagnetic bcc structure is observed, while the indentation modulus and hardness values are found to be 160 and 8.7 GPa, respectively. For x = 0.49, a strongly $\langle 200\rangle$ -textured paramagnetic phase is evidenced, having the A15 structure. Mössbauer measurements also allowed the site occupancy to be determined. The structural change from bcc to A15 phase yields a 50% increase in hardness, while the indentation modulus remains unchanged. This particular behaviour is discussed in terms of distances between near neighbours and Burgers vector length in both bcc and A15 structures. The results obtained here provide motivation to pursue the present study through wear experiments.

The novel VCP system is a mobile physical deposition method to deposit metallic/magnetic films using various source materials including powder, lump, pre-alloyed ingots and wires. The VCP system consists of a large deposition area of 960 cm2 and has been used for the first time to prepare magnetic thin films of Si3Fe97. The source material evaporated by a resistively heated furnace, which was position right under the substrate within the VCP system, contains small pieces of conventional 3% silicon-iron steel as source materials. The magnetic analysis of the films was achieved by using a vibrating sample magnetometer (VSM). Observations indicate that the magnetic anisotropy and coercivity are dependent on the type of substrate and the deposition conditions. Results of all films deposited on flexible kaptonTM are anisotropic in the film plane whereas the films deposited on glass substrate indicate the less-well defined anisotropy in the film plane while the substrate holder of the VCP system was run at the speed of 100 rpm. In the case of stationary magnetic materials production, the films deposited on kapton and glass substrates show isotropic magnetic behaviour. All films showed planar magnetic anisotropy irrespective of type of substrate and the production conditions used. The findings are discussed in terms of scaling up the technique for the possible production of various shapes of circular, square or strip components with the compositions equivalent to that of conventional electrical steels in order to investigate a possible future to produce large scale of silicon-iron as the core materials for rotating machines and power transformers.

The pick-up coil of a conventional flux-gate sensor has beenreplaced by a magnetic tunnel junction. With this respect, themagnetic properties of tunnel junctions have been optimized to berelevant to flux-gate operation i.e. low-coercitive field andreproducibility of the dynamic switchings. Furthermore, the use ofan alternative two-dimensional magnetic driving field extends theuse of the flux-gate like sensor to make a 2D magnetometer with asingle magnetic tunnel junction without any cross sensitivity. Afirst demonstration is given with sensitivity above 500 V/T anda noise level evaluated to 2  $\rm \mu T/ \sqrt{Hz}$ with an optimizedmagnetoresistive response.

Cathode-directed sparks engrave in a specially prepared, originally low-conducting nonlinear silver-based material beyond a critical electric power input highly structured discharge patterns which exhibit an anomalous potential distribution consisting of an electrical double layer in series with a region of high electrical conductivity.

Measurements of electrostatic force generated in thedrift region of a positive unipolar wire-to-plane corona dischargein air are compared with numerical simulations. The generated forcedoes not depend on the ion mobilities, preventing the influence ofuncertainty and variation of these parameters. A method is describedto simulate the voltage and charge distribution for a wire-to-planeset-up. This method enables the determination of the transitionbetween unipolar and bipolar discharges. In the experimental set-upbreakdown electric field of air reduces with increasing dischargecurrent.

In this paper the measurements of the dynamic breakdownvoltages U b for linearly rising pulses in nitrogen at lowpressure are presented. The measurements were carried out for therates of voltage rise k up to $300\, {\rm V\,s}^{-1}$ . Dependence of thebreakdown voltages, delay times and electron yields on the rate ofrise were obtained experimentally and theoretically under different conditions. It was found thatthe overvoltage $\overline {\Delta U_b}$ and the mean effectiveelectron yield $\overline {YP}$ is proportional to $\sqrt k$ (Yis a number of generated electrons in the interelectrode space persecond and P the breakdown probability), while the statisticaltime delay $\overline t_s$ is proportional to $1/\sqrt k$ . In thesecond part, the experimental breakdown voltage distributions wereobtained, fitted by theoretical distributions and some breakdownparameters relevant to experimental conditions were determined.Based on the approximate analytical and numerical models, thedependence of the effective secondary electron yield γ onthe overvoltage and on the rate of voltage rise were derived fromthese measurements. It was found that γ varies linearlywith the overvoltage for a constant k, and the slope of γis proportional to  $\sqrt k$ .