Table of contents

X-ray lithography (XRL) is a patterning technique for the fabrication of semiconductor devices with very small dimensions (⩽100 nm). In this process the device and circuit layout images are formed by exposing a thin layer of organic material, the photoresist. XRL is based on the idea that the short wavelength (≈1 nm) of the radiation allows a faithful reproduction of the desired pattern to very small dimensions. The physics of the interaction of the x-rays with matter dictates many aspects of the technique: the imaging process is controlled by Fresnel diffraction, and the recording process by the interaction of low-energy photoelectrons with the resist material. Phase shift effects play a key role in the image formation process, and are used to improve the sharpness of the image. To achieve high-resolution patterning it is necessary to carefully design the pattern itself, and optimize several exposure parameters. This paper presents a detailed discussion of the image formation process, and a brief review of other technical aspects of the technology.

A study on the transient heat and mass transfer from vertical plates embedded in fluid saturated porous media is conducted. The wall temperature and concentration are power functions of the dimensionless streamwise coordinate. Results for the heat and mass transfer rates are obtained for different values of the Lewis number, the buoyancy ratio, the power-law exponent and the ratio of porosity and heat capacity ratio. The transient local Nusselt number at X = 1 is found to be almost independent of the ratio of porosity and heat capacity ratio, while the transient local Sherwood number at X = 1 tends to increase as the ratio of porosity and heat capacity ratio is increased. Moreover, increasing the power-law exponent increases the transient local Nusselt and Sherwood numbers.

The present article describes an experimental study of the dynamics of a small magnetic particle moving on the flat horizontal free surface of a rotating liquid and submitted to a magnetic field that decreases with distance from a dipole magnet. The particle is kept at the surface by interface forces and its motion is bidimensional. For a constant external force, its behaviour displays a transition at a threshold angular rotation velocity Ω = Ω_c.

Below the threshold, the particle remains at a fixed point; above the threshold, its trajectory tends towards a limit cycle of ovoidal shape and size increasing with Ω.

A phenomenological modification of the equation that describes the motion of the particle is suggested in order to represent the influence of capillary effects on the particle trajectory: an effective drag coefficient is introduced that decreases the inertial response time of the particle.

The numerical integration of the equation agrees with experimental observations. Nonlinear effects for a time-dependent magnetic field have been studied in the case of a square wave variation of frequency ω. The shape of the Poincaré sections is shown to depend on the ratio ω/Ω.

Starting from the disc angular coordinate, a circle map may be constructed which shows an almost one-dimensional behaviour.

X-ray diffraction (XRD) and magnetization measurements of Y₂(Fe_1-xCo_x)₁₅Ga₂ (0⩽x⩽1) compounds were performed. All samples were single-phase with a hexagonal Th₂Ni₁₇-type structure. The substitution of Co for Fe leads to a nearly linear decrease in the unit cell volume and a monotonic increase of Curie temperature. The average magnetic moment of the 3d metal atom increases with x and then decreases, reaching a maximum at x = 0.2. The magnetic anisotropy phase diagrams of Y₂(Fe_1-xCo_x)₁₅Ga₂ (0⩽x⩽1) compounds have been determined from the temperature dependence of magnetization and the XRD patterns of magnetically oriented powder samples. The spin-reorientation temperature was found to increase with increasing Co content for 0.4<x<1.

Magnetostrictive FeCo films (thickness 300 nm) have been produced by RF magnetron sputter deposition. The effects of the substrate, composition and thermal treatment on the structural and magnetic properties of the films have been determined. Structural analysis was performed using transmission electron microscopy and x-ray diffraction. The magnetic properties, including magnetostriction, were determined by the magneto-optical Kerr effect, magnetic force microscopy and strain-modulated ferromagnetic resonance. It is found that the magnetic softness of the films is critically dependent on the texture and strain state of the film. With suitable choices of substrate, composition and thermal treatment, these parameters can be controlled, producing magnetically soft films while maintaining a high magnetostriction. The differential response of the magnetic anisotropy to strain in these films is comparable to the best values achieved by more involved manufacturing processes, such as multilayering, showing excellent potential for their use in magnetic sensors and actuators.

Granular CoCrPt-SiO₂ thin films with Cr underlayers have been fabricated by sputtering followed by post-deposition annealing. Microstructural and magnetic properties of as-deposited and annealed films have been investigated for potential applications in magnetic recording media. The grain size of about 10 nm was obtained in the as-deposited films, and subsequent annealing did not lead to significant grain growth. High coercivity up to 5640 Oe was achieved upon annealing. The M_rt value can be down to about 0.4 memu cm^-2 for films with H_c exceeding 5000 Oe. The peak value of the delta M curves changed from a positive value in as-deposited films to a negative value for annealed films, indicating an obvious reduction of intergrain magnetic exchange coupling after annealing.

The structure evolution of granular (NiFeCo)_xAg_(1-x) (x = 9-41 at%) films was characterized by x-ray photoelectron spectroscopy, Rutherford backscattering spectroscopy, x-ray diffraction, atomic force microscopy and magnetic force microscopy. The giant magnetoresistance of the films was measured as a function of temperature between 20 and 300 K using a conventional four-point probe dc technique in the presence of a magnetic field up to 7.6 kOe. The temperature dependence of magnetization and magnetic hysteresis loops for the films were measured by a SQUID magnetometer. It was found that the optimum concentration and annealing temperature for the maximum giant magnetoresistance was associated with the crystalline structure and the magnetic domain structure of the film. A clear flat-top parabola and a significant deviation from the quadratic law expected for equal-size, non-interacting superparamagnetic particles in the magnetoresistance (Δρ/ρ) against magnetization (M/M_s) curve were observed for the 500 °C annealed (NiFeCo)₂₀Ag₈₀ sample in a wide field region. The curves of Δρ/ρ against M/M_s were well described by a function of the form c(M/M_s)¹⁰. This behaviour was explained by combining the characteristics of the microstructure, magnetic domain structure and magnetic properties of the sample.

In the geometric-optics limit the electromagnetic (EM) continuity equation, which describes EM energy flow, reduces to the radiation transport equation, in which energy flow is described generically as the motion of light (or particle) rays along straight-line trajectories in the direction of a unit vector . In practical applications is discretized, making the transport equation extremely difficult to implement computationally and giving rise to `ray effect' distortions owing to the non-continuous distribution of the energy over . In this paper we solve the EM continuity equation directly in the transport regime, obtaining an energy which is continuously distributed in and free of ray distortion.

The numerical study presented in this paper investigates the role of thermal lensing on the mode matching between the pump beam and the cavity beam in continuous-wave end-pumped Cr⁴⁺:forsterite lasers. A suitable mode overlap function was first derived to calculate the degree of overlap between the pump beam and the fundamental mode of the resonator. The effect of thermal lensing on mode matching was then numerically investigated by calculating the changes in the average value of the mode overlap function due to variations in pump power and crystal boundary temperature. Thermal lensing was taken into account by calculating the pump-induced thermal gradients and by approximating the gain medium as a distributed quadratic lens. Finally, the model was used to offer guidelines about how one of the resonator lenses should be readjusted in order to maintain optimum mode matching as the boundary temperature and pump power are varied.

LIF spectroscopy has been introduced for the in situ diagnostic of phenol degradation in the aqueous solution under pulsed-corona discharges. The applicability of LIF to the measurements of phenol concentration up to 1×10^-5 mol l^-1 was confirmed. The temporal variation of the absolute concentration of phenol was measured in an aqueous solution exposed by pulsed-corona discharges. This exhibited two characteristic decay times.

We discuss the properties of a dc discharge which occurs at the interfaces between a disc of fused glass beads and parallel electrodes. The discharge consisted of a large number of microdischarges. The discharge current had a large number of pulses superimposed on a dc background current. The background current increased with decreasing bead size and decreased as the separation between the disc and the electrode increased. The current had an extremely slow and nonlinear transient behaviour when the dc high-voltage source was switched on. The frequency of the current pulses decreased with increasing bead size and disc-electrode separation. The average amplitude of the pulses, in contrast, increased with increasing bead size and disc-electrode separation. A significant dependence of the properties of the current pulses on the polarity of the applied voltage was observed, although the dc current was almost independent of the polarity. It was found that the enhancement of the electric field near the contact points between the electrode and the beads, where the current concentrates, initiates the discharge. The density of ozone produced by the discharge was also measured. Damage to the discs by the discharge was also observed, thereby altering the properties of the discharge.

This study was dedicated to the correlation between the rotational temperature and the plasma gas temperature in a non-thermal atmospheric plasma obtained by a dielectric barrier discharge. Different parameters were modified (e.g. gas temperature, molecular fraction) in argon water gas mixtures. The rotational temperature of OH(A) was spectroscopically measured. It is shown that in mixtures mainly composed of atomic species, namely rare gases, the OH(A) rotational temperature is strongly correlated with a gas temperature streamer. In mixed (atomic and molecular) plasmas it seems that the rotational temperature can constitute an approach to the translational energy of heavy particles in the streamer areas.

The reaction of SiH₄ with nitrogen is studied in a post-discharge device by means of a mass spectrometer. The atomic nitrogen is produced in a dc discharge and the silane is injected downstream and reacts mostly with the atomic nitrogen. Induced effects due to the vibrational excitation of the molecules (change in the ionization cross section when molecules are ionized in the ionization chamber of the mass spectrometer) are minimized by adding a buffer gas (N₂) that is injected downstream with the silane.

The reaction scheme of the silane injected in the nitrogen post-discharge, proceed via the dissociation of the silane with atomic nitrogen according to SiH₄ + N⟶^k₁SiH₂ + NH₂. Then the two radicals SiH₂ and NH₂ produced react also with the atomic nitrogen via the reactions SiH₂ + N⟶^k₂products and NH₂ + N⟶^k₂products. These two radicals react also together via SiH₂ + NH₂⟶^k₄products.

The absolute concentrations of the main species considered in the reaction mechanism are measured using different methods and absolute concentrations of the species are compared with results given by a kinetic model. A Runge-Kutta-Merson algorithm is used in order to resolve the set of ordinary differential equations. The experimental results are well fitted with a relative accuracy of 20%, using the reaction rate constants k₁, k₂, k₃ and k₄ in the ranges (1.2-1.7)×10^-16 m³ s^-1, (9.0-13.7)×10^-16 m³ s^-1, (5.8-8.8)×10^-16 m³ s^-1 and (4.2-6.2)×10^-16 m³ s^-1, respectively.

Xe dielectric barrier discharges at different gap lengths under applied pulse voltages with trapezoidal and sinusoidal waveforms were simulated using a self-consistent one-dimensional fluid model. In both waveforms, the light output power depended not only on the amplitude of voltage waveforms but also on the discharge gap length. At the narrower discharge gap, the light output efficiency was improved by increasing the time gradient of the applied voltage when the trapezoidal pulse is applied, and by decreasing the duty ratio in the sinusoidal case. In the present simulation, we adopted a fast numerical method for calculation of electric field introducing an exact expression of the discharge current.

The results of two-dimensional modelling of positive streamer dynamics in 2 cm point-to-plane gap are presented, including the process of photoionization in a gas volume. The model of photoionization in air, developed by Zhelezniak, Mnatsakanian and Sizykh is considered in detail and a comparison with the experimental data of Penney and Hummert is presented. The model of photoionization in nitrogen, based on the data of Penney and Hummert is described. The simulations of streamers were performed for air and nitrogen-like gas. It is shown, that fundamental space scale of streamer - the width of space charge layer l_ρ (ionization domain) around its head is defined by the length of absorption of photoionizing radiation l_a. This length defines all other parameters: electron density, radius of streamer channel etc, so one may speak about standard streamer in air, whose properties depend only on pressure. Based on simulation results one may assume that streamer branching occurs in high field, when l_ρ exceeds l_a. The process of branching may be interpreted as an instability which transforms the non-standard streamer into a number of standard ones.

Measurements of the complex elastic Young's modulus of PZT ceramics as a function of the frequency of a triangular switching electric field and of an ac dynamic mechanical stress were performed. The results show that the electrostrictive character of the ceramics provides a complex elastic Young's modulus, which depends on the 90° domains reorientation velocity which, in turn, is driven by the low-frequency switching electric field. This dependence has a relaxational character with a τ = 30 s relaxation time. For a given frequency of the switching electric field, the behaviour of Young's modulus as a function of the frequency of the mechanical stress is quite similar to that corresponding to point defects that are pinning the ferroelectric 90° immobile domain walls: it is relaxational with a relaxation time τ = 0.04 s.

Two experimental arrangements have been used to measure the thermal diffusivity at room temperature using laser pulse heating. One uses circular disc specimens and the other rectangular plate-like specimens to measure both one-dimensional and three-dimensional heat transfer. In both experimental configurations a pulsed laser beam which has a circular cross section is irradiated onto the front surface of the specimen. The time dependence of the temperature response of a circular area on the back surface of the specimen is measured using a radiation detector. The experimental data is then used to calculate thermal diffusivity using appropriate analytical expressions. The theory of these methods is outlined and results are discussed. In addition, finite element models have been generated for one-dimensional and two-dimensional transient thermal analysis of the specimens. These results are shown to be in close agreement with the results of experiments within the bounds of experimental error, hence, validating the experimental measurements.

Silicon carbide is an important wide-band-gap semiconductor for high temperature, high-voltage, high-power and high-frequency devices. Electrical isolation is an important aspect for device applications. In this report, oxygen ions, 70 keV with doses ranging from 5×10¹³ to 5×10¹⁵ cm^-2, have been implanted into n-type 6H-SiC to investigate the possibility of forming a high-resistive layer. The damage behaviour and internal stress were checked by Rutherford backscattering spectroscopy and channelling, and an x-ray rocking curve, respectively. Atomic force microscope observations revealed that the surface morphology is quite sensitive to the implantation even at a dose of 1×10¹⁴ cm^-2. After annealing in nitrogen at 1200 °C, no remarkable damage recovery could be seen if the deposit damage energy is over the critical value. Schottky structures of Au/SiC have been fabricated on the annealed samples and I-V curves of metal/SiC/InGeNi were measured at room temperature at both forward and reverse bias; the electrical isolation effect was observed at proper implantation dosages. The results indicated that there exists a dose window for electrical isolation.

A simple method for characterizing synthetic graphite powders is presented in this paper. A thick glass cylinder contains a known weight of material, which is held between two conducting pistons. These allow the measurement of the electrical conductivity of the sample, and enable its compaction. The graphite powder is gently pressed, and its volume and conductivity variation are simultaneously measured. From these results, the relation between conductivity and the volume fraction of grains is derived: an equation based on the effective medium theory (EMT) is shown to fit the experimental data accurately. The adjustable parameters are directly linked with both the anisotropic conductivity and the morphology of the grains. On the one hand, conductivity measurements achieved on single particles give values of the same order of magnitude as those derived from the fits. On the other hand, the other parameters of the equation perfectly agree with the aspect ratios obtained from apparent density measurements. The percolation thresholds which would be expected from composites made of graphite powders imbedded in an insulating medium are also calculated via the EMT equation. Comparison with other theoretical and experimental values always leads to very good agreement, showing the accuracy of the EMT equation in supplying fair geometrical parameters for the particles.