Table of contents

Metallic elements having negative enthalpies of mixing tend to form characteristic local atomic clusters. In this review, we use the structural information in the first nearest neighbour shell level, or first-shell atomic cluster, to derive the composition rules of two types of complex alloy phases, quasicrystals and bulk metallic glasses, both being composed of elements with negative enthalpies of mixing. We first show the composition phenomena in quasicrystal-forming systems, where major composition rules such as cluster line, electron concentration and atomic size criteria are derived. Then we analyse the composition rules of bulk metallic glasses using the very same approaches. Finally, we summarize their common composition rules into more general rules and basic theories.

The magnetic behaviour of Sm(Co,Fe,Cu,Zr)_z ribbons with z = 7.0 and 7.5 is investigated. The Henkel plots indicate that there is a strong intergrain exchange coupling (IGEC) effect in these ribbons. It is the IGEC that results in the phenomena of the interaction domains and remanence enhancement in the ribbons at room temperature. The contributions of IGEC and dipolar interaction to irreversible magnetization have been analysed to explain the phenomenon of open recoil loops. Due to the difference of the intensity of the IGEC, the magnetization changes with the component of the Sm(Co,Fe,Cu,Zr)_z ribbons.

This paper presents a systematic method to design and calculate tunnelling magneto-resistance (TMR) sensors with high signal-to-noise ratio (SNR). The sensing module consists of four TMR devices arranged in a Wheatstone-bridge configuration. Closed-form equations were obtained to calculate TMR sensor current, array output voltage, magneto-resistance ratio, overall noise (thermal and shot) and SNR for a given bandwidth. Using this technique we were able to maximize the SNR by tuning the many parameters of the TMR devices. Typical SNR values are in excess of 45 dB.

A number of valved quartz glass ³He spin filter cells have been repeatedly exposed to various external magnetic fields in order to determine the influence of induced wall magnetization on the relaxation time in the cells. The procedures of magnetizing and degaussing of cells are described. A comparison of T₁ measurements performed in the same cell by different methods attest the good reliability of the measurements as well as the time stability of T₁ in Cs-coated quartz glass cells. No orientation dependence of the relaxation in fields of 8 G was observed. A strong dependence of T₁ on the strengths of external magnetic fields, applied perpendicular to the direction of the guide field used during the measurements, was observed at low and moderate fields. T₁ decreases rapidly by increasing the field strength up to values of about 1 kG both in Cs-coated and bare-wall cells. Further increase in the field to 7 kG has no significant influence on the relaxation value, suggesting the saturation of the cell magnetization. Macroscopic magnetization produced in the bulk of the glass as well as Cs-caused magnetization on the surface are proposed to be two different relaxation mechanisms present in the cells. The magnetization induced in cells has been determined independently of the ³He relaxation by SQUID measurements. The influence of Cs coating and of heating on T₁ in magnetized cells is discussed.

For the first time the magnetocaloric properties of La_0.9Ag_0.1MnO₃, La_0.8Ag_0.2MnO₃, La_0.85Ag_0.15MnO₃, La_0.8Ag_0.15MnO₃ and La_0.8Ag_0.1MnO₃ manganites have been investigated by direct and indirect measurement techniques. All samples showed almost the same relative cooling power (RCP). Temperatures of maxima of the magnetocaloric effect (MCE) are between a few degrees below freezing and the room temperature region. The compounds showed RCP values of about 100 J kg⁻¹ at a field change of 2.6 T, which is about half the RCP of gadolinium. Because of considerable MCE and the Curie temperatures ranging from 269 to 303 K, these materials could be used as magnetic refrigerants for magnetic refrigeration in the sub-room and room temperature range.

Based on the target-field method of solving Fredholm integral equations of the first kind, a new approach is presented in the paper for designing gradient coils that can be used in a permanent-magnet magnetic resonance imaging (MRI) system with biplanar poles. To restrict the current distribution on the coil plane within a finite radius, the current density is pre-expanded into Fourier series by orthogonal basis functions. By setting the target-field points and B_z values over the imaging region of interest, corresponding integral equations are derived from the Biot–Savart law to calculate the current densities. They form a matrix equation, in which the unknown elements of the column vector are the Fourier coefficients for the unknown current density. As long as these target-field points are well chosen, the Fourier coefficients can be solved by inverse matrix calculation instead of the regularization method for Fredholm integral equations of the first kind. Then the current density is discretized using the stream-function method to generate the winding patterns. To verify the feasibility of this approach, the gradient magnetic field generated by the current density is calculated via the Biot–Savart law. Optimized parameters are obtained through computer simulations for some shielded and unshielded transverse gradient coils. The performance of this approach has been demonstrated as well.

Fe–MgO nanocomposites synthesized by the mechanical high-energy transfer technique were characterized by x-ray diffraction (XRD), transmission electron microscopy, Mössbauer spectroscopy and dc magnetization studies. By varying the duration of milling, powder with different grain sizes in the range 17–40 nm was produced. XRD and Mössbauer measurements could not detect the presence of any form of iron oxide in the nanocomposites. Transmission electron micrographs showed a shape transformation from spherical to acicular-like geometry in the samples ball milled for more than 36 h. Dc magnetization measurements show an increase in the coercive field with decrease in average grain size, with a sharp increase at the point of shape transformation. A marked decrease from the expected values of saturation magnetization and magnetic moment per formula unit was observed. This decrease was ascribed to the magnetic dead layer formed at the surface of the nanoparticles, which increased with the increase in milling time. The nanocomposites ball milled for more than 36 h exhibit a two-component Mössbauer spectra due to the presence of two phases: a crystalline phase and an additional phase corresponding to the atoms residing at the grain boundaries between the crystallites. The grain boundaries give rise to a distribution of hyperfine magnetic fields, typical of an amorphous-like compound.

Giant magnetoimpedance (GMI) was studied in electrodeposited Co/Cu/Co-coated Ag wires with variations in the axial dc magnetic field (0–2 kOe), frequency of ac driving current (1 kHz–100 MHz), thickness of each layer (1.0–25.0 µm) and diameter of the Ag core (47.7 and 120.0 µm). With increasing thickness of coated layers and Ag cores, the maximum GMI ratio increased but the characteristic frequency decreased. Unlike Co-coated Ag wires, the frequency-dependent GMI curves of these multilayered wires were broadened and asymmetric. The GMI behaviours were explained in terms of the distribution of current density and current flow in the coated wires obtained from the finite element method simulation program.

In this work, we fabricated white organic light-emitting devices (WOLEDs) containing a layered light-emitting region composed of a single blue-emitting host and different fluorescent dopant materials. The effects of varying the dye-doping ratio and emitting layer thickness on the efficiency, lifetime, spectral voltage-dependence and white balance were investigated for devices with a blue/orange stacked layer structure. Addition of a blue host layer doped with a green-emitting dopant, to give a blue/green/orange emitter, resulted in a broadband white spectrum without the need for a charge-blocking interlayer. The composition of blue, green and orange dopants in the host and the thickness of each emitting layer were optimized, resulting in a device efficiency of 9–11 cd A⁻¹ even at a high brightness of 10 000 cd m⁻² (achieved at a bias voltage of less than 9 V) with an emission spectrum suitable for lighting applications.

This paper reports the fabrication and measurement of organic light-emitting devices comprised of indium-tin-oxide (ITO)/4,4',4''-tris(N-carbazolyl)-triphenylamine (TCTA, 8 nm)/2-pyrenyl-9-phenyl-9-pyrenylfluorene(2P9PPF, 30 nm)/ bathocuproine (BCP, with different thickness)/Alq₃ : DCJTB (DCJTB = 4-(dicyanomethylene)-2-t-butyl-6(1,1,7,7- tetramethyljulolidyl-9-enyl)-4H-pyran, Alq₃ = tris(8-hydroxyquinolino) aluminium(III), 2% in mass ratio, 30 nm)/Mg : Ag(250 nm). The dependence of electroluminescence (EL) spectra on the BCP layer thickness and operating voltage has been investigated quantitatively. It is shown that the emission colour of the devices changes from red to blue at 8 V when the BCP layer thickness changes from 1 to 10 nm. The emission colour of the devices also varies with the applying voltage even in a given device. The ratio of blue emission originating from 2P9PPF to the red emission originating from DCJTB increases with the applying voltage. Based on the hole blocking effect of the BCP layer, we deduced the dependence of this ratio on the BCP layer thickness and simulated the experimental result well. It is also proposed that the variation of the EL spectrum with the voltage can be attributed to the varying hole blocking effect under the varying electric field, which resulted in a recombination zone shift, and the exciton dissociation effect in the electric field.

The pulsed laser deposition of YBa₂Cu₃O_7−x targets by excimer laser at fluences of 4–10 J cm⁻² in low pressure oxygen backgrounds yields emissive plumes with kinetic energies of 50–200 eV, driving the formation of a shock front with Mach numbers of M = 10–50. The propagation of the shock front is independent of atomic species and adequately characterized by the Sedov–Taylor shock model if the dimensionality of the plume is allowed to deviate from ideal spherical expansion. The ideal efficiency of energy conversion from laser pulse to shock expansion is nearly unity at 1 Torr, but decreases rapidly at lower pressures, where the plume expands beyond the laser footprint during ablation. The low oxygen background pressures, 100–1000 mTorr, typically employed for the production of superconducting films is sufficient for the generation of a strong shock front with shock thickness of 5 mm to less than 0.4 mm, but too low to develop three-dimensional flow. Indeed, dimensionality of the expansion ranges from n = 0.8 to 2.4 over the background oxygen pressure range of 25–1000 mTorr. Shock strength is proportional to the Mach number and inversely dependent on pressure, indicating a thickness limited to approximately the collision mean free path.

Proton-exchange LiNbO₃ planar optical waveguides have been irradiated with swift ions (Cl 30 MeV) at very low fluences in the range 5 × 10¹⁰−5 × 10¹² cm⁻². Large modifications in the refractive index profiles, and therefore in the optical performance, have been obtained due to the generation of amorphous nano-tracks by the individual ion impacts. Moreover, a fine tuning of the refractive index can be achieved by a suitable control of the fluence (δn/δϕ ∼ 10⁻¹⁴ cm² or δn ≈ 10⁻⁵ for δϕ = 10⁹ cm⁻²). An effective medium approach has been used to account for those changes and determine the amorphous fraction of material. The results have been compared with information extracted from complementary RBS channelling experiments. From the calculated amorphous fractions a radius of ∼2 nm for the amorphous tracks have been estimated.

The present paper reports a study of optical parameters, refractive index (n) and extinction coefficient (k), by analysing the transmission spectrum of thin films prepared from samples of Ge₂₀Se_80−xIn_x (x = 0, 5, 10, 15, 20) in the range 400–1500 nm. The dispersion parameters are calculated by using the Wemple–DiDomenico model. The optical band gap of the thin films is obtained by Tauc plots. The variation in the optical parameters for different thin films is explained on the basis of the presence of defect states and the change in the average bond energy of the system. The dielectric constants (ε_r and ε_i) and optical conductivity (σ) of the thin films are also reported.

We have investigated the effect of thin fluoro-acrylic polymer as an anode stabilizer on the lifetime of an organic light emitting device (OLED). Surface chemical properties of commercial fluoropolymer, FC-722 (Fluorad™ of 3M), on indium–tin oxide (ITO) were characterized by x-ray photoemission spectroscopy. An OLED with 1 nm thick fluoropolymeric film showed identical brightness and efficiency behaviour and improved operational stability compared with the reference device with UV-O₃ treated ITO. The improvement in the lifetime was accompanied by the suppression of the voltage increase at the initial stage of constant-current driving, which can be attributed to the action of the FC-722 layer by smoothing the ITO surface. Fluoropolymer coating, therefore, improves the lifetime of the small molecular OLED by the simple and reliable anode-stabilizing process.

We have developed and characterized diethynyl naphthalene derivatives with high ionization potentials, 2,6-bis [(naphthalen-2-yl)ethynyl]-naphthalene (BH-202) and 2,6-bis[(heptafluoronaphthalen-2-yl)ethynyl]-naphthalene (BF-202) for new organic semiconductors. Using x-ray diffraction, we have found that thin films of the two organic semiconductors have a high degree of molecular order. Organic field-effect transistors (OFETs) with BH-202 films showed p-type characteristics. By the optimization of the fabrication process, the device exhibited a high field-effect hole mobility up to 0.12 cm² V⁻¹ s⁻¹ and a high on/off current ratio of 3.3 × 10⁵. On the other hand, OFETs with BF-202 films showed n-type characteristics. The field-effect electron mobility was calculated to be 3.4 × 10⁻³ cm² V⁻¹ s⁻¹ and the on/off current ratio was 5.4 × 10³.

A single-electrode type interdigital transducer (SIDT) is extensively used to excite surface acoustic waves (SAWs) on piezoelectric substrates because of its simple structure and fabrication process. In a SIDT a bus-bar is used for feeding in electrical voltage. However, systematic analyses about bus-bar effects of SIDT are still sparse among the existing studies concerning the propagation of acoustic wave of SIDT. In this paper, a two-dimensional interface Green's function method is adopted for calculating the electrical charge density on the electrode surfaces and acoustic wave fields in a Y-cut Z-propagation LiNbO₃ substrate excited by SIDTs with a bus-bar. The computation results show that the charge density associated with SAW cannot be neglected; the bus-bar has essential effects on the excited SAW fields and the SAW field generally diverges strongly due to the influence of the bus-bar. Approaches for reducing this effect are discussed.

A functional structure material with good thermal insulation properties, which can also reflect infrared light within a certain range, was obtained by introducing a photonic crystal structure as a reflector into oxidized porous silicon. The porous silicon multilayer was fabricated based on alternative changes in etching current intensity. The oxidation process was performed to ensure a full oxidation of the high porosity layers and a partial oxidization of the low porosity layers to achieve good thermal insulation properties while maintaining enough contrast of the refractory index. The microstructure was characterized by scanning electron microscopy and the optical reflectance spectra by Fourier transform infrared spectroscopy. A band gap centred at 3 µm with a bandwidth of 1.3 µm was successfully obtained.

The influence of evaporation conditions of organic semiconductor material tris(8-hydroxyquinoline)aluminium (Alq₃) on the performance of organic light emitting diodes (OLEDs) is reported. In the process of organic material thermal evaporation, the chamber was dark or had white light from a 100 W filament lamp. The devices prepared in the dark show higher emission intensity and efficiency compared with the ones prepared in white light under the same driving voltage. Atomic force microscopy measurements show that surface morphology and phase of Alq₃ thin films are quite different for the previous cases. The different evaporation conditions are found to have direct effects on the electrical and luminance performance. The Alq₃ thin films prepared in the dark as active emission layers of OLEDs show higher intensity and efficiency.

We report on the infrared laser-induced photovoltaic effect of a heterostructure comprising a n-ZnO film with an electron concentration of 1.6 × 10²¹ cm⁻³ and a MgB₂ film with a high carrier concentration of 1.8 × 10²³ cm⁻³ at room temperature. Current–voltage characteristics of the structures have good rectifying diode-like behaviour. An open-circuit photovoltage of 10 ns rise time and 20 ns full width at half-maximum was observed under the illumination of a 1.064 µm pulsed laser for a duration of 25 ps. When the junction was irradiated by a 10.6 µm CO₂ laser pulse for 60 ns, a transient photovoltaic signal of ∼27 mV occurred with a rise time of ∼60 ns. A possible mechanism of the observed phenomena is discussed based on the laser-induced hot carrier photovoltaic effect.

Plasma decay after a high-voltage nanosecond discharge was studied experimentally and numerically in room temperature N₂, CO₂ and H₂O for pressures between 1 and 10 Torr. The time-resolved electron density was measured by a microwave interferometer for initial electron densities in the range 8 × 10¹¹–3 × 10¹² cm⁻³ and the effective electron–ion recombination coefficient was determined. It was shown that this coefficient varies in time and depends on pressure. A numerical simulation was carried out to describe the temporal evolution of the densities of charged particles under the conditions considered. A good agreement was obtained between the calculated and the measured electron density histories. It was shown that the loss of electrons is governed by dissociative recombination with complex ions, their density being dependent on pressure. In N₂ at low pressures, a hindered electron thermalization in collisions with molecules led to a delay in the plasma decay. This effect was observed both experimentally and theoretically.

The firing voltage (FV) of gas discharge in a test cell of plasma display materials was investigated for standard protective layers of MgO deposited by electron beam, and ternary BaY₂O₄ and BaGa₂O₄ oxides grown by pulsed laser deposition on the special dielectric coated glass substrates. The determined FVs for MgO (160 V), BaY₂O₄ (210 V) and BaGa₂O₄ (257 V) lead to the conclusion that a replacement of MgO by these ternary oxides is not expedient for plasma display panels. Using results from luminescence spectroscopy, values for the energy gap E_g ≈ 6.2 and 5.8 eV were estimated for BaY₂O₄ and BaGa₂O₄, respectively. The main reason for the observed high FVs is attributed to strong electron affinities χ, where χ (BaY₂O₄) < χ (BaGa₂O₄).

A planar inductively coupled radio-frequency (rf) magnetic neutral loop discharge has been designed. It provides diagnostic access to both the main plasma production region as well as a remote plane for applications. Three coaxial coils are arranged to generate a specially designed inhomogeneous magnetic field structure with vanishing field along a ring in the discharge—the so-called neutral loop (NL). The plasma is generated by applying an oscillating rf electric field along the NL, induced through a four-turn, planar antenna operated at 13.56 MHz. Electron density and temperature measurements are performed under various parameter variations. Collisionless electron heating in the NL region allows plasma operation at comparatively low pressures, down to 10⁻² Pa, with a degree of ionization in the order of several per cent. Conventional plasma operation in inductive mode without applying the magnetic field is less efficient, in particular in the low pressure regime where the plasma cannot be sustained without magnetic fields.

The absolute number density of oxygen atoms in the O(³P₂) state is determined in an inductively coupled plasma chamber using absorption spectroscopy enhanced by a high finesse cavity with a continuous wave diode laser as the radiation source. At a nominal pressure of 100 mTorr of precursor gas O₂ and 500 W plasma power the O(³P₂) concentration is found to be (4.8 ± 0.6) × 10¹⁴ cm⁻³.

Comparison of different Monte Carlo codes for understanding their limitations is essential to avoid systematic errors in the simulation, and to suggest further improvement for the codes. MCNP4C and EGSnrc, two Monte Carlo codes commonly used in medical physics, were compared and evaluated against electron depth-dose data and experimental results obtained using clinical radiotherapy beams. Different physical models and algorithms used in the codes give significantly different depth-dose curves. The default version of MCNP4C calculates electron depth-dose curves which are too much penetrating. The MCNP4C results agree better with the experiment if the Integrated Tiger Series-style energy-indexing algorithm is used. EGSnrc uses a class II-Condensed History (CH) scheme for the simulation of electron transport. To conclude the comparison, a timing study was performed. It was noted that EGSnrc is generally faster than MCNP4C and the use of a large number of scoring voxels dramatically slows down the MCNP4C calculation. However, the use of a large number of geometry voxels in MCNP4C only slightly affects the speed of the calculation.

Results of the simulation of transient plasma discharges in a stoichiometric atmospheric-pressure methane–air mixture are presented. It is shown that current control allows one to produce non-thermal plasma with parameters (the gas temperature, the molar fractions of active species) varying over a wide range. The calculated plasma parameters are used in the modelling of ignition of the mixture. Estimated values of the ignition delay time and the minimum ignition energy are given.

Extensive electron collision data for Ar and Xe are assembled and employed to study the e-beam deposition in this noble gas mixture for application to the Ar–Xe laser. Nineteen states of Xe and thirteen of Ar are used for the atomic model. The electron energy distribution function is calculated from the Boltzmann equation using atomic cross sections for excitations from the ground states computed from atomic physics codes. The resulting electron distribution is then used to investigate the electron temperature, energy per electron–ion pair, ionization rates and excitation-to-ionization ratios for both species as a function of three parameters: the Xe mole fraction, the fractional ionization, and the power deposition per Ar atom. While the Ar ionization and excitation are fairly insensitive to these parameters, the rates for Xe, especially total excitation-to-ionization ratio, can vary with all three. Cross sections for excited to excited transitions among all the considered upper states in Xe are also calculated and transition rates, fitted to an Arrhenius form, are presented in tabular form. The assembled atomic data and calculated rates are useful for further detailed electron kinetics modelling of the Ar–Xe laser.

In order to study the short pre-arcing time in high breaking capacity (HBC) fuses, we use a mathematical model including the phase change of the fuse-element heating coupled with the Laplace equation for the potential and Ohm's law. The thermal model is based on the enthalpy formulation of the heat equation with a source term representing the Joule heating. For the time range considered (up to 10 ms), we assume no heat transfer between the fuse-element and the surrounding sand. To solve numerically the governing equations, we employ a semi-implicit scheme for time integration and a finite element method for space discretization. Using electrical and thermal properties of the silver fuse-element, we present pre-arcing characteristics (temperature, current density, potential) for a fuse-element used in industrial protection circuits.

The development and propagation of Trichel pulses in oxygen have been numerically simulated using an improved fluid particle-in-cell (PIC) method. The numerical method has been optimized to accurately reproduce sequences of about 100 Trichel pulses (∼1 ms). A classical one-dimensional model of negative corona in sphere-to-plane geometry has been used to formulate the continuity equations for electrons and ions. The effects of ionization, attachment and secondary-electron emission from the cathode have all been considered. The electric field has been obtained from the solution of Poisson's equation in two dimensions. Using this model, the temporal and electrical characteristics of Trichel pulses have been investigated, in particular, the relation between applied voltage, pulse frequency and time-averaged current intensity and charge.

Experimental investigations of the transformer-coupled toroidal discharge (TCTD) in a mixture of argon and mercury vapour were carried out for a wide range of discharge parameters: mercury vapour pressure (0.1–40 000 Pa), discharge current (1–260 A), diameter of a gas-discharge tube (20–75 mm) and current frequency (10–250 kHz). New experimental data on electric field strength and emission yield to visible and UV-range were obtained as functions of the discharge parameters. It was proved that the dependences of electrical and optical characteristics of TCTD on the discharge current, vapour pressure and tube diameter are close to those of dc arc discharges in mercury vapour. The possible reasons for parameter deviations were considered. Also, the effect of the reduction of emissivity for the visible triplet of mercury spectrum (404.7, 435.8, 546.1 nm) with the excitation potential of 7.73 eV was discovered under certain conditions: mercury vapour pressure is higher than 10 kPa, and the specific power is above 50 W cm⁻¹.

Irradiation of thin films of sub stoichiometric germanium oxide (GeO_x) has been carried out using swift heavy ions. Micro-Raman studies of the films indicate the formation of Ge crystallites as a result of irradiation. Moreover, crystallinity of Ge improves with an increase in fluence. Glancing angle x-ray diffraction results also confirm the presence of Ge crystallites in the irradiated samples. Fourier transformed infrared spectroscopy was employed to study the Ge–O–Ge vibrational structure before and after irradiation. Formation of Ge crystallites has been explained on the basis of the phenomenon of ion-beam-induced phase separation.

This paper reports nanoscale piezoelectric measurements on ferroelectric poly(vinylidene fluoride-co-trifluoroethylene) (P(VDF-TrFE)) films prepared by the Langmuir–Blodgett (LB) technique. Polarization patterning, piezoelectric hysteresis and relaxation after poling are studied in this work by piezoresponse force microscopy. High quality P(VDF-TrFE) films with a copolymer content of 30% were fabricated using a Schaefer monolayer transfer setup permitting precise control of the film microstructure. The thickness of the films deposited with 100 transfers was ∼64 nm. Local switching resulted in written polarization lines with the lateral size in the range 70–300 nm depending on polarization time. Local hysteresis loop (at a fixed tip position) demonstrated clear ferroelectric switching with the coercive voltage ≈8–10 V that corresponds to a macroscopic switching field (∼1.5 MV cm⁻¹) at a ∼10 nm depth below the tip. Relatively slow ageing after poling was observed with the characteristic relaxation time of about 1500–2000 s depending on the polarization direction. The obtained results demonstrate that the stable polarization patterns can be created in LB P(VDF-TrFE) films and attest them as suitable candidates for memory and nanotemplate applications.

We report the fabrication of YBa₂Cu₃O_7−δ/Nb-doped (0.8 wt%) SrTiO₃ p–n junctions by pulsed electron deposition, photolithography and Ar-ion etching techniques. In the junctions, the abnormal rectifying behaviour and the Esaki tunnelling diode-like characteristic have been observed above and below the superconducting transition temperature T_C of YBa₂Cu₃O_7−δ, respectively. The tunnelling process is responsible for these behaviours and the depletion layer formed at the junction interface with very low resistance is the dominating factor. The intriguing properties in this work indicate that it is possible to realize the functions of traditional semiconductor diodes by using high-temperature superconductor-based heterojunctions.

The influence of annealing and of wire geometry on the giant magnetoimpedance (GMI) effect in a glass-coated microwire LC-resonator were systematically investigated. It was found that annealing magnetic microwires significantly varied the permeability and hence the GMI ratio, whereas the magnetoimpedance response was sensitively altered with varying wire geometry of the LC-resonator. The magnitude of GMI is determined by the soft magnetic characteristics of the microwire and the effect of LC-resonance, while the multiple-peak GMI feature can be attributed to the LC-resonance circuit and the formation of standing magnetic waves within the sample. Interestingly, the GMI ratio reached the largest values of 400 000% and 270 000% at the resonance frequencies of 518.51 MHz and 146.32 MHz for Co_83.2B_3.3Si_5.9Mn_7.6 and Co₆₇Fe_3.8Ni_1.4B_11.5Si_14.6Mo_1.7 samples, respectively. The corresponding magnetic-field sensitivities of GMI are about 218 000%/Oe and 135 000%/Oe. These results are interesting for the development of a new family of ultra-sensitive and high-frequency magnetic sensors. The feature of a LC-resonance circuit can be used for improving the sensitivity of GMI-based magnetic sensors while selecting the working frequency.

The magnetotransport properties of La_0.75Sr_0.25MnO₃ (LSMO) films epitaxially grown on SrTiO₃ (0 0 1) substrate were studied. The Curie temperature (T_c) of LSMO films decreases from 300 to 105 K with a reduction in the film thickness to 5 nm. A clear metal–insulator transition is observed at a temperature consistent with T_c for films thicker than 10 nm. However, the 5 nm thick film displays an insulative characteristic along with a high magnetoresistence effect in a wide range of temperatures below and above T_c. A phase-separation phenomenon in the 5 nm thick film was demonstrated by thermally activated hopping transport between ferromagnetic metallic domains embedded in an insulative matrix at temperatures below T_c, but between paramagnetic metallic domains at temperatures above T_c. It was also confirmed by conductive atomic force microscopy images.

High density arrays of Zn_1−xCo_xO (x = 0.05, 0.10, 0.15, denoted by ZnCoO) nanorods were vertically grown on glass coated ZnO films via hydrothermal reaction at 70 °. The structures and morphology of the arrays were studied by x-ray diffraction, photoluminescence (PL) spectroscopy and field emission scanning electron microscopy, which show that the nanorods of 150 nm diameter and 4.5 nm length grow along the [0 0 0 1] direction. X-ray photoemission spectroscopy demonstrated that Co was successfully doped into the nanorods. Ammonia concentration and ZnO buffer were found to play a very important role in the nucleation and growth of the ZnCoO nanorods. PL spectra were composed of broad ultraviolet (UV) emission and visible light. Using Gaussian simulation, the UV band was separated into two peaks, both of which were characteristic of redshift with the increase in Co concentration. The effect of the Co dopant on the broad ultraviolet emission peak and the formation mechanism of ZnCoO arrays were also discussed.

Gas phase nitridation of nc-TiO₂ was carried out for obtaining nc-TiN. The nc-TiO₂ precursor itself was derived from a Ti(OH)_x precipitate formed through hydrolysis of titanium nitrate solution by aqueous ammonia reagent. Calcination of titanium hydroxide at 1000 °C yielded nc-TiO₂ powder with an average particle size of ∼5 nm. Both nc-Ti(OH)_x as well as nc-TiO₂ were separately nitrided in an environment of argon and ammonia for synthesizing nc-TiN. A home-built thermogravimetric analysis-mass spectrometer apparatus was used for the nitridation experiment. The nanocrystalline phases of Ti(OH)_x, TiO₂ and TiN were characterized by XRD, TEM, IR and XPS techniques. The powders' XRD patterns indicated the formation of nanocrystalline TiN by ammonia bases nitridation of nc-TiO₂. These studies also revealed sequential crystallo-chemical transformation of the intermediate precursor. FTIR studies revealed the existence of IR active modes in nc-TiO₂ and the formation of a chemically pure TiN phase. The TEM image of TiN powder depicted the formation of nanocrystalline phase. Ti(2p_3/2) photoelectron spectra indicated complete conversion of nc-TiO₂ to nc-TiN.

Thermal degradation of GaP(0 0 1) surfaces has been studied using spectroellipsometry (SE), optical microscopy, scanning electron microscopy (SEM), ex situ atomic force microscopy (AFM), x-ray photoelectron spectroscopy (XPS) and photoluminescence (PL) techniques. The SE data suggest that thermal annealing causes removal of the native oxide at temperatures T < 600 °C and surface roughening at T ≥ 600 °C. The XPS spectra confirm the native oxide removal from the GaP surface. The AFM images provide rms roughnesses of ∼0.6 nm at T < 600 °C and ∼2 nm (∼15 nm) at T = 600 °C (800 °C), suggesting a rapid increase in the rms roughness with an increase in T above 600 °C. The reduction in PL intensity is observed for T ≥ 500 °C. The Ga droplets, observed by optical microscopy and SEM, appear at T > 600 °C, i.e. after the native oxide is removed from the GaP surface. These facts allow us to draw the conclusion that the surface native oxide acts as a good passivation film against annealing-induced PL degradation and thermal decomposition.

We have investigated the growth of CdTe thin films on glass substrates by hot wall epitaxy. The layers have been characterized by scanning electron microscopy, atomic force microscopy, profilometry, x-ray diffraction and optical transmission. The grown samples are polycrystalline with a high preferential [1 1 1] orientation. Atomic force microscopy and scanning electron microscopy reveal pyramidal grain shapes with a size of around 0.3 µm. The surface roughness increases with sample thickness and growth temperature, reaching about 200 nm for 10 µm thick layers grown at 300 °C. Samples with a thickness of 2 µm grown at 150 °C showed a roughness of less than 40 nm. Optical transmission measurements demonstrate layers with high optical quality.

The present study is devoted to co-deposited Ag : BN nanocermet thin films and is focused on the influence of ion irradiation conditions on their structural and linear optical properties. Ion irradiation was performed in situ during the growth of the nanocermets using a 50 eV assistance beam (nitrogen/argon or nitrogen-ion assistance) and ex situ on as-grown films using a 120 keV argon-ion beam (post-irradiation). Grazing incidence small-angle x-ray scattering measurements show that (i) as-grown N-assisted films contain prolate spheroidal clusters (height-to-diameter ratio H/D ≈ 1.8), (ii) N/Ar-ion assistance leads to the formation of more elongated clusters (H/D ≈ 2.1) and (iii) post-irradiation leads to a decrease of H/D to a value close to 1. These results are discussed on the basis of atomic diffusion processes involved during the growth of the nanocermets and during the post-irradiation. The optical transmittance spectra of these films measured at normal incidence display one absorption band, due to the excitation of the (1,1) plasmon mode of the clusters. In the case of the as-grown films, an additional band appears at oblique incidence for P-polarized light, as a consequence of the excitation of the (1,0) plasmon mode of the clusters. Our results show that the spectral position of the absorption bands (which can be tuned in the 400–600 nm range) depends on the H/D ratio of the clusters, in good agreement with calculations of optical transmittance considering the nanocomposite layer as a uniaxial anisotropic medium whose dielectric tensor is described by an anisotropic Maxwell–Garnett model.

Hollow zinc oxide microspheres have been successfully synthesized on silicon substrate via simple thermal evaporation of pure zinc powder without any catalyst or template material at a relatively low temperature of 550 °C. The morphologies, chemical composition and crystal structure are characterized using x-ray powder diffraction, transmission electron microscopy and scanning electron microscopy. A smooth surface and rough ones with many nanopores are observed. The difference in the structures of the shells is discussed and a possible growth mechanism is proposed in detail. The room-temperature photoluminescence spectrum reveals a peak at 380 nm corresponding to free exciton emission and a strong green emission at ∼ 525 nm associated with defect-related emission.

High-density structures with well-ordered nanohole arrays have been obtained by the self-ordering growth of nanopores using 25–40 V anodization voltages in a sulfuric–oxalic acid mixture. In each anodization voltage, two types of electrolyte mixture were used with low and high sulfuric concentrations at a constant pH value. The ordering and regularity of a nanohole array was quantified by analysis of fast Fourier transformations of scanning electron and atomic force microscope images. The results show that the long range ordering of nanohole array is superior for high sulfuric concentration, while regularity is more pronounced for low sulfuric concentration. The interpore distance in the oxalic–sulfuric mixture is higher in the case of low sulfuric concentration. It is almost proportional to 2.6 nm V⁻¹ and 2.5 nm V⁻¹ for the electrolyte mixture with low and high sulfuric concentrations, respectively. While the porosity is nearly 10% for samples made in standard sulfuric and oxalic acid using 25 V and 40 V anodization voltage, respectively, it increases up to 50% for the sample made with 37.5 V in an oxalic–sulfuric mixture.

The paper describes a magnetoelectrical technique for monitoring the particle velocity in matter acquiring a large electrical conductivity under shock compression. The particle velocity is measured in the electromagnetic skin layer, which travels through matter with a wave velocity. The layer thickness essentially depends on the electrical conductivity of shock-compressed matter. A multielectrode cell also allows the wave velocity to be registered at various spatial bases. Experiments give kinematics parameters of a metallization wave for aluminium powders of varied mass density and particle size. The results unambiguously testify that the conductivity of the aluminium powder arises in the main shock wave carrying full pressure. The results do not confirm the conclusion of Novac et al about metallization of the aluminium powder in an elastic precursor. The thickness of the skin layer for a coarse powder is smaller than the thickness of the shock-compression zone, which offers the opportunity of finding the metallization phase directly in the shock transition zone.

Phase change materials (PCMs) are used for efficient thermal energy storage. When a PCM melts and solidifies, it absorbs and releases a large amount of heat within a small temperature interval. Salt hydrates are interesting PCMs with high storage density, but their solidification is often problematic due to large subcooling. From thermodynamic theory, it should be possible to cause nucleation by applying high pressure to the subcooled melt, and thereby reduce subcooling. However, for the design of a pressure based triggering system there are still many unknown factors. In this context, we investigated the pressure dependence of the melting and nucleation temperatures. We present experimental data of three inorganic PCMs under static pressures up to 800 MPa. For NaOAc · 3H₂O we observed a shifting of the nucleation temperature from −20°C at ambient pressure to +40°C at 800 MPa. This confirms that within this pressure range, the nucleation temperature of NaOAc · 3H₂O is shifted above room temperature. For CaCl₂ · 6H₂O, a good agreement with reported melting temperature data was observed, and the range of experimental data was extended. For KF · 4H₂O, the shift of the melting temperature was found to differ considerably from theoretic predictions.

We describe new experiments on the sequential imbibition (capillary absorption) of the immiscible liquids n-decane and dimethylsulphoxide into dry concrete. A Sharp Front model is developed and experimental data are analysed in terms of this model. The model predicts a long-term asymptotic t^1/2 behaviour and a new hybrid sorptivity property of the porous medium.

Crystal structural parameters of Bi_4−xNd_xTi₃O₁₂ (BNT, x = 0.25, 0.5, 0.75 and 1.0) were determined by means of neutron powder diffraction and Rietveld analysis based on the monoclinic space group of B1a1. The a-axis has been found to shrink monotonically from 5.4475(8) to 5.4100(1) Å with x increasing from 0 to 0.75. The orthorhombicity defined as 2(a−b)/(a+b) decreased with increasing x because the b-axis lattice parameter remained almost constant. The c-axis increases monotonically up to x = 0.75. The spontaneous polarization, which is oriented along the a-axis (P_s//a), decreased from 35.4 to 28.3, 24.8 and 22.2 µC cm⁻² with x increasing from 0 to 0.25, 0.5 and 0.75, respectively as calculated by using the refined structural parameters. Thus, the substitution of Nd for Bi has been found to cause a considerable structural relaxation. Bi₃Nd₁Ti₃O₁₂ showed a considerably small a-lattice parameter of 5.4004 (1) Å and a very small P_s value of only 7.5 µC cm⁻² indicating that its structure is orthorhombic or close to tetragonal.

InAlGaN/GaN heterostructures with various Al compositions have been grown on sapphire substrate using the metal organic chemical vapour deposition technique. The solid-to-gas phase ratio indicates a high Al incorporation efficiency. Atomic force microscopy reveals a smooth surface with the formation of hexagonal pits. The size and the density of the hexagonal pits increase with increasing Al mole fraction. The Hall effect and the capacitance–voltage (C–V) studies show the formation of a two-dimensional electron gas (2DEG) at the InAlGaN/GaN interface. A relatively higher Hall sheet carrier density compared with the 2DEG density estimated from the C–V profile indicates parallel conduction via the underlying GaN layer. It is observed that the 2DEG density decreases as a function of the Al composition and these results are discussed based on the increasing depth of the hexagonal pit and the background donor density.

Elastomer compounds reinforced with precipitated silica can exhibit elevated wet sliding friction on a rough surface in comparison with corresponding compounds filled with carbon black particles. The underlying mechanism is currently not well understood. To unravel this puzzling observation, the variation of wet sliding friction with filler volume fraction is examined at the sliding speed of the order of 1 m s⁻¹ under different lubrication conditions. Depending on the lubrication liquid—water or ethanol—a compound that shows both higher bulk hysteretic loss and lower modulus does not always exhibit a higher wet sliding friction. A thorough characterization of the bulk rheology of the compounds investigated fails to provide the rationale for such behaviour, thus constituting an apparent violation of the conventional viscoelastic understanding of rubber friction on a rough surface. On the other hand, the detected lowering of friction when the lubrication liquid is changed from water to ethanol resembles the effect of liquid medium on interfacial adhesion reported in the literature. Hence, it is suggested that a stronger interfacial attractive interaction should exist in water between the road surface and silica particles on the compound surface immediately next to the road surface. This should be related to the elevated wet sliding friction detected for silica-filled compounds under water lubrication.

Dielectric and ferroelectric properties of chemically synthesized and sintered lead lanthanum zirconate titanate and lead zirconate titanate discs were studied. The as-burnt powders were directly sintered to 94% and 96% of the theoretical density at 1100 °C and 1200 °C, respectively for lead lanthanumm zirconate titanate (PLZT) and 96.7% at 1100 °C for lead zirconate titanate (PZT). The sintered PLZT samples showed improved dielectric properties whereas PZT showed improved ferroelectric properties compared with the data reported in the literature. Dielectric permittivity at room temperature of the 1100 °C sintered PLZT sample with grain sizes 0.4–1.2 µm was ∼5000, at all measured frequencies. Evaporation of PbO from the sample generated micro inhomogeneity, which was detrimental for the dielectric property. A sintered PLZT sample underwent a completely diffused phase transition whereas a sintered PZT sample underwent a partially diffused phase transition. The measured values of remnant polarization of the 1100 °C sintered PZT and PLZT samples were 54.0 and 40.0 µC cm⁻², respectively. The coercive fields observed for these samples were 13.3 and 7.7 kV cm⁻¹, respectively.

Glasses in the system Na_2−2xB_4−4xBi_xMo_0.5xO_7−4x, 0 ≤ x ≤ 0.4, have been prepared by the melt quenching technique. Elastic properties and IR spectroscopic studies have been employed to study the role of Bi₂O₃ and MoO₃ on the structure of Na₂B₄O₇ glass. Elastic properties and Debye temperature have been investigated using sound velocity measurements at 4 MHz. The results showed that the density and the molar volume increase while both sound velocities and the determined glass transition temperatures decrease with increase in x. Infrared spectra of the glasses reveal that the strong borate network consists of diborate units and is affected by the increase in the concentration of Bi₂O₃, and MoO₃. These results are interpreted in terms of the increase in the number of non-bridging oxygen atoms, substitution of longer bond lengths of Bi–O, and Mo–O in place of shorter B–O bond and the change in Na⁺ ion concentration. The results indicate that bismuth and molybdenum ions have been substituted for boron ions as network modifier ions. The elastic moduli are observed to increase with the increase in Bi₂O₃ and MoO₃ content. This contradiction in the elastic moduli–molar volume relation is attributed to the role of the respective bonds.

We investigate p-type doping in ZnO prepared by metal-organic chemical vapour deposition with dimethylhydrazine (DMHy) as the nitrogen dopant source. Results obtained by x-ray photoelectron spectroscopy show that DMHy exhibits a narrow temperature window from 500 to 550 °C for efficient nitrogen incorporation and that nitrogen doping is critically influenced by growth conditions, e.g. the N/Ga flux ratio in growth. Within an appropriate N/Ga flux ratio range, p-type ZnO can be realized. The effect of the N/Ga flux ratio on the conductivity conversion of ZnO is reported. The extrinsic nitrogen acceptor level is calculated to be about 160 meV from low-temperature photoluminescence spectra.

An experimental system of a four-pad 3ω-method with noise cancellation circuit was designed to measure simultaneously thermal conductivity, thermal diffusivity and heat capacity of an individually suspended wire. The closed solution of temperature oscillation, in which the heat transport in the wire is modelled as one-dimensional diffusion in vacuum, was developed in the frequency domain for fitting the experimental results. Both the closed solution and the series solution were used to determine the thermal properties. The measurements on an individual polyacrylonitrile-based carbon fibre 7 µm in diameter were performed using both the closed solution and the series solution, which yielded the thermal conductivity of 84.4 W m⁻¹ K⁻¹, the heat capacity of 1.23 MJ m⁻³ K⁻¹ and the thermal diffusivity of 7.16 × 10⁻⁵ m² s⁻¹ at room temperature. The relative uncertainties of the measurements were analysed.

The shear strengths of three lead-free solder joints have been measured over the range of loading rates 10⁻³ to ∼10⁵ mm min⁻¹. Binary (SnAg), ternary (SnAgCu) and quaternary (Castin: SnAgCuSb) alloys have been compared to a conventional binary SnPb solder alloy. Results show that at loading rates from 10⁻³ to 10² mm min⁻¹, all four materials exhibit a linear relationship between the shear strength and the loading rate when the data are plotted on a log–log plot. At the highest loading rate of 10⁵ mm min⁻¹, the strengths of the binary alloys were in agreement with extrapolations made from the lower loading rate data. In contrast, the strengths of the higher order alloys were found to be significantly lower than those predicted by extrapolation. This is explained by a change in failure mechanism on the part of the higher order alloys. Similar behaviour was found in measurements of the tensile strengths of solder joints using a novel high-rate loading tensile test. Optical and electron microscopy were used to examine the microstructures of interest in conjunction with energy dispersive x-ray analysis for elemental identification. The effect of artificial aging and reflow of the solder joints is also reported.

Zr-substituted BaTiO₃ thin films (BaTi_1−xZr_xO₃: BTZ) were prepared by rf-sputtering on SrTiO₃ (STO) substrates or STO substrates buffered by pulsed laser deposited SrRuO₃ (SRO) electrodes. The epitaxial growth of BTZ was monitored in situ by reflective high energy electron diffraction. X-ray reflectivity measurement demonstrated that a smooth surface with a roughness of 0.5 nm could be achieved under optimized deposition parameters. Film composition and the change in chemical environment due to Zr substitution are discussed based on electron energy-loss spectroscopy. The high quality of the interfaces was confirmed by transmission electron microscopy. The epitaxial relation between BTZ/SRO and SRO/STO was studied by selected area diffraction and high resolution transmission electron microscopy. All the results show that high quality epitaxial BTZ thin films were successfully deposited and ready for further electrical studies.

Highly (1 1 1)-oriented (Pb,Ba)ZrO₃ (PBZ) thin films were deposited on BaPbO3(BPO) electrode at different deposition temperatures and different thicknesses. As the deposition temperature increased, the dielectric constant and tunability increased from 150 and 4% to 500 and 56%, respectively. A current transient peak observed prior to dielectric degradation can be related to oxygen vacancy migration (redistribution) with space-charge-limited current transient or modulation of the electric conductivity. We infer that an interfacial layer lying between the PBZ and BPO effectively increases the oxygen vacancy (density ∼10²⁰ cm⁻³) mobility by Pb⁴⁺ diffusing into the PBZ films.

Investigations on ion transport behaviour of new Ag⁺ ion conducting quaternary solid electrolyte systems: x[0.75AgI : 0.25AgCl] : (1 − x)RbI, where 0.1 ≤ x ≤ 1 mol.wt fractions, are reported. The solid electrolyte systems were synthesized adopting different routes of solid solution reactions. An alternative host salt, 'a quenched [0.75AgI : 0.25AgCl] mixed system solid solution', discovered at the present laboratory, has been used in place of AgI. The compositional dependent conductivity studies at room temperature revealed the quaternary systems 0.7[0.75AgI : 0.25AgCl] : 0.3RbI as the optimum conducting composition (OCC). AgI has been traditionally and most widely used as host salt in the past to prepare the majority of well-known fast Ag⁺ ion conductors including the ternary superionic solids MAg₄I₅ (M = Rb, K, NH₄). Also, RbAg₄I₅ has been recognized as one of the highest conducting superionic solids with conductivity σ_rt ∼ 10⁻¹ S cm⁻¹ at room temperature. However, it has been observed that RbAg₄I₅ is thermodynamically unstable, particularly in the open ambient conditions, while the newly synthesized quaternary solid electrolyte, having OCC 0.7[0.75AgI : 0.25AgCl] : 0.3RbI, has been observed to be much more stable than RbAg₄I₅. For direct comparison of the ion conduction behaviour of the newly synthesized OCC, the ternary superionic solid RbAg₄I₅ (0.8AgI : 0.2RbI) has also been synthesized in an identical manner. The room temperature conductivity (σ_rt) of the newly synthesized quaternary OCC was slightly lower than that of RbAg₄I₅. However, σ_rt remained practically stable for a fairly long time in the open ambient conditions as compared with that of RbAg₄I₅. The mechanism of ion transport in OCC has been characterized on the basis of experimental studies on various ionic parameters namely conductivity (σ), ionic mobility (μ), mobile ion concentration (n), ionic drift velocity (v_d), ionic transference number (t_ion) etc. The temperature dependent studies on these ionic parameters provided information regarding various energies involved in the respective thermally activated processes. Solid state batteries have been fabricated using OCC as electrolyte in the cell configuration Ag-metal (anode)/0.7[0.75AgI : 0.25AgCl] : 0.3RbI (solid electrolyte)/C+I₂ (cathode) and the cell potential discharge characteristics have been studied under various load conditions.

The tremendous heat generated in a computer chip or very large scale integrated circuit raises many challenging issues to be solved. Recently, liquid metal with a low melting point was established as the most conductive coolant for efficiently cooling the computer chip. Here, by making full use of the double merits of the liquid metal, i.e. superior heat transfer performance and electromagnetically drivable ability, we demonstrate for the first time the liquid-cooling concept for the thermal management of a computer chip using waste heat to power the thermoelectric generator (TEG) and thus the flow of the liquid metal. Such a device consumes no external net energy, which warrants it a self-supporting and completely silent liquid-cooling module. Experiments on devices driven by one or two stage TEGs indicate that a dramatic temperature drop on the simulating chip has been realized without the aid of any fans. The higher the heat load, the larger will be the temperature decrease caused by the cooling device. Further, the two TEGs will generate a larger current if a copper plate is sandwiched between them to enhance heat dissipation there. This new method is expected to be significant in future thermal management of a desk or notebook computer, where both efficient cooling and extremely low energy consumption are of major concern.

This work involves a study on laser ignition (LI) in an internal combustion (IC) engine and investigates the effects on control of engine combustion performance and stability of varying specific laser parameters (beam energy, beam quality, minimum beam waist size, focal point volume and focal length). A Q-switched Nd : YAG laser operating at the fundamental wavelength 1064 nm was successfully used to ignite homogeneous stoichiometric gasoline and air mixtures in one cylinder of a 1.6 litre IC test engine, where the remaining three cylinders used conventional electrical spark ignition (SI). A direct comparison between LI and conventional SI is presented in terms of changes in coefficient of variability in indicated mean effective pressure (COV_IMEP) and the variance in the peak cylinder pressure position (Var_PPP). The laser was individually operated in three different modes by changing the diameter of the cavity aperture, where the results show that for specific parameters, LI performed better than SI in terms of combustion performance and stability. Minimum ignition energies for misfire free combustion ranging from 4 to 28 mJ were obtained for various optical and laser configurations and were compared with the equivalent minimum optical breakdown energies in air.

Grid-shaped and tree-shaped flow architectures are being generated for use in vascularized materials with multiple functionality (self-healing, self-cooling, etc), based on the principle of the constructal law of evolutionary increase of flow access through the generation of better flowing configurations (designs). Here we investigate systematically the advantages of endowing the complex flow architecture with more freedom to morph. Four ways to increase design freedom are explored: multi-scale grids (one, two, three and four diameter sizes), multi-shape loops (square, triangular, hexagonal, rhombic), multi-shape bodies (hexagonal, square, rhombic) and vascularization with grids versus trees. We show that significant gains in global flow access are achieved as the number of optimized channel diameters increases. The most promising combinations of body and loop shapes are hexagonal bodies with triangular loops and square bodies with square loops. The tree-shaped architecture outperforms the grid, but it is recommendable only for stressed bodies in which the most likely location of the cracks is known ahead of time. The effect of body size on the global performance of vascularized multi-scale and multi-shape materials is documented. Diminishing returns and increasing robustness set in as the complexity of the optimized flow architectures increases.

On page 4127, in the fifth line of section 4, Discussion, the text `The value of 4.64 eV for ϕ of the CNT in [6] ...' should read as `The value of 4.64 eV for ϕ in the present paper ...' This remark was intended to show that the results in [12] agree closely with the result for ϕ calculated in the present paper.