Table of contents

The transformation of Pd/Si to Pd₂Si/Si is investigated using depth-resolved positron annihilation, x-ray diffraction and Auger electron spectroscopy studies. The observed defect-sensitive positron S-parameter value of 1.022–1.054 indicates the existence of divacancies across the silicide/silicon interface and Si substrate region. Our experimental observation of vacancy defects is consistent with the model proposed for excess vacancy generation across the interface consequent to Si diffusion.

We have carried out de Haas–van Alphen (dHvA) experiments on a ferromagnet CeRh₃B₂ with an extremely high Curie temperature  K and a non-4f reference compound LaRh₃B₂. The dHvA data of LaRh₃B₂ are well explained by the results of energy band calculations. The topology of the Fermi surfaces in CeRh₃B₂ is found to be very similar to that of LaRh₃B₂, possessing wavy but flat Fermi surfaces in the basal plane. Observation of a quasi-one-dimensional electronic state is the first such case in a rare earth compound.

We have investigated the spin-gap in high T_c superconductivity. We obtain the effective exchange integral in the presence of conduction in the ab-plane from the interaction U_sd, where the electron–electron interaction is mediated by the localized spin flips. We choose the exchange interaction along the c-axis from the superexchange-type interaction, U_sd^c. We find the spin-gap from the conducting spin- ladder corresponding to the structure of high T_c superconductors.

The structure and electrical activity of monatomic hydrogen defect centres are inferred from the spectroscopy and charge-state transitions of muonium, the light pseudo-isotope of hydrogen. Introductions are given to all these topics. Special attention is paid to the shallow-donor behaviour recently established in a number of II–VI compounds and one III nitride. This contrasts with trapped-atom states suggestive of an acceptor function in other members of the II–VI family as well as with the deep-level amphoteric behaviour which has long been known in the elemental group-IV semiconductors and certain III–V compounds. The systematics of this remarkable shallow-to-deep instability are examined in terms of simple chemical considerations, as well as current theoretical and computational models. The muonium data appear to confirm predictions that the switch from shallow to deep behaviour is governed primarily by the depth of the conduction-band minimum below the vacuum continuum. The threshold electron affinity is around 3.5 eV, which compares favourably with computational estimates of a so-called pinning level for hydrogen (+/−) charge-state transitions of between −3 and −4.5 eV. A purely ionic model gives some intuitive understanding of this behaviour as well as the invariance of the threshold. Another current description applies equally to covalent materials and relates the threshold to the origin of the electrochemical scale. At the present level of approximation, zero-point energy corrections to the transition levels are small, so that muonium data should provide a reliable guide to the behaviour of hydrogen. Muonium spectroscopy proves to be more sensitive to the (0/+) donor level than to the (+/−) pinning level but, as a tool which does not rely on favourable hydrogen solubility, it looks set to test further predictions of these models in a large number of other materials, notably oxides. Certain candidate thin-film insulators and high-permittivity gate dielectrics appear to be uncomfortably close to conditions in which hydrogen impurity may cause electronic conduction.

Geometric quantum computation is a scheme to use non-Abelian holonomic operations rather than the conventional dynamic operations to manipulate quantum states for quantum information processing. Here we propose a geometric quantum computation scheme which can be realized with current technology on nanoscale Josephson junction networks, known as a promising candidate for solid-state quantum computers.

The x-ray multiple diffraction method, which allows us to determine the piezoelectric coefficients of a single crystal under an external electric field (E), was applied to Rochelle salt (dos Santos et al 2001 J. Phys.: Condens. Matter13 10497) for E parallel to the piezoelectric Y direction. In this work, the theory was extended to consider an observed monoclinic–triclinic distortion under E application into the other two piezoelectric X and Z directions. Renninger scans carried out using the chosen (060) primary reflection have provided the four remaining coefficients through the measurement of the properly chosen secondary peaks. so that all eight piezoelectric coefficients (d₁₄, d₁₆, d₂₁, d₂₂, d₂₃, d₂₅, d₃₄ and d₃₆) for Rochelle salt were determined.

The influence of both compositional changes and ordered/disordered structure on hyperfine parameters of the body-centred-cubic α-FeCo phase is discussed in terms of the binomial distribution and nearest-neighbour approximation. In the nanocrystalline alloy series Fe_78−xCo_xNb₆B₁₅Cu₁, it was found that, for the alloy with x = 18, nanocrystalline grains present a disordered structure, whereas for the alloys with 39 and 60 at.% Co, a tendency to an FeCo atomic ordering is observed.

A selection of the vibrational modes of two variants, compact and non-compact, of three-membered self-interstitial clusters in crystalline Si is studied by first-principles calculations in order to contribute to the possible identification of a three-bodied cluster with the W centre in Si. Internal vibrational modes of the compact clusters are found between 19 meV ('translation' modes) and 47 meV (breathing mode), in fair agreement with the phonon replica spectrum of W. However, the compact cluster does not possess the local mode above the bulk phonon range associated with W. A local mode is instead exhibited by the non-compact cluster variant. The latter, though, is energetically disfavoured over the compact one. The combined results on the vibration and energetics suggest that neither of the two clusters can be identified with W.

The validity of a perturbative approach to studying the lifetime of excited electrons in an electron gas is analysed in a range of metallic densities. The relaxation rate is calculated using a kinetic theory framework with second-order Born and partial-wave approximations for the scattering amplitude. The comparison of the terms obtained by physically motivated Thomas–Fermi-type potentials shows that the next-to-leading correction to the first-order result is small only for high densities of the electron gas. At low densities, a nonperturbative estimation of the relaxation rate is needed. An analytic expression of practical use which incorporates both the role of screening and scattering is derived. A comparison with experimental data obtained for an Ag(111) surface is made. The agreement found supports an inherent consistency, under the experimental conditions and down to (E−E_F) = 0.4 eV, of the theoretical description based on a three-dimensional Fermi-liquid model.

We study the optical conductivity of the one-band Hubbard model in the Néel state at half filling at T = 0 using the dynamical mean-field theory. For small values of the Coulomb parameter clear signatures of a Slater insulator expected from a weak-coupling theory are found, while the strongly correlated system can be well described in terms of a Mott–Heisenberg picture. However, in contrast to the paramagnet, we do not find any evidence for a transition between these two limiting cases but rather a smooth crossover as a function of the Coulomb interaction.

First-principles density functional calculations on the La₂CuO₄ crystal, simulated by using the Cu₅O₂₆/Cu₈La₃₄ cluster, have been analysed to reveal that the Cu 4s orbital is occupied by about 0.5 electrons. Since this may have important consequences on the method of calculation of the intrinsic hole distribution in cuprates, a study of the frontier orbitals has been made. It is concluded that the Cu 4s occupancy is a direct result of a charge transfer from the oxygen anions but does not involve the hole. It is a clear illustration that the hole distribution cannot always be estimated from the charge density distribution alone. 60% of the hole remains on the copper while the rest is spread evenly about the planar oxygen atoms.

The first-principles calculation within density-functional theory is used to search for new candidates of half-metallic ferromagnets in semi-Heusler alloys NiCrM (M = P, As, Sb, S, Se and Te). Our calculations predict that NiCrP, NiCrSe and NiCrTe are half-metallic ferromagnets (HMFs) with magnetic moments of nearly 3 or 4 μ_B/fu and HM gaps of 0.263, 0.047 and 0.102 eV, respectively. Other compounds are so-called nearly HMFs. Substitution of the sp atoms cannot be responsible for the formation of the band gap, but results in a shift in the Fermi level and the loss of half-metallicity. The half-metallicity of NiCrP and NiCrTe can be retained when the lattice parameter is changed by about 2%–3%.

In pseudo-tetragonal strontium bismuth tantalate, SrBi₂Ta₂O₉ (SBT), with two formula units per unit cell, bismuth oxide {(Bi₂O₂)²⁺} layers alternate with double strontium tantalate perovskite layers {(SrTa₂O₇)²⁻}. A unit cell of SBT is truncated to form a sub-cell or film, of composition SrBi₂Ta₂O₁₁, which is 1.4 nm thick and comprised of a bottom (BiO₂)¹⁺ layer, a central (SrTa₂O₇)²⁻ layer and a top (BiO₂)¹⁺ layer. Using spin-polarized first-principles calculations, it is found that this SrBi₂Ta₂O₁₁ film is multi-ferroic, magnetoelectric, i.e. it simultaneously exhibits ferroelectric and ferromagnetic characteristics. When Ta ions are collectively displaced in the ab plane and in the [110] direction, the calculated double potential energy well, with a depth of −3.1 eV/unit cell at a Ta off-centre displacement of 0.032 nm, reflects the ferroelectric character. The calculated spin-polarized electronic structure reveals that ferromagnetism stems, not from the d electrons of the Ta ions, but predominantly from the unpaired p electrons of the O ions. The O ions in the Sr–O layer have the largest magnetic moment of 1.32 μ_B. Specifically, the ferromagnetic character is mediated by the unoccupied states of the Sr 5p band above the Fermi level, E_F. These states provide a mechanism for the double exchange or hopping of highly localized O 2p (majority) spins between adjacent O ions located on both sides of the Sr ion.

We propose a superlattice model to describe superconductivity in layered materials, such as the borocarbide families with the chemical formulæ RT₂B₂C and RTBC, with R being (essentially) a rare earth, and T a transition metal. We assume a single band in which electrons feel a local attractive interaction (negative Hubbard-U) on sites representing the TB layers, while U = 0 on sites representing the RC layers; the multi-band structure is taken into account minimally through a band offset ε. The one-dimensional model is studied numerically through the calculation of the charge gap, the Drude weight, and the pairing correlation function. A comparison with the available information on the nature of the electronic ground state (metallic or superconducting) indicates that the model provides a systematic parametrization of the whole borocarbide family.

A theoretical model is suggested which describes a non-conventional relaxation mechanism in strained high transition temperature superconducting films, namely the formation of nanograins with 90° grain boundaries. It is theoretically revealed here that misfit stresses in superconducting thin films at early stages of their growth are effectively relaxed through the formation of nanograins with their crystal lattice misoriented by 90° relative to the crystal lattice of the film matrix. With increasing film thickness, the formation of a continuous layer resulting from the convergence of nanograins becomes energetically favourable. The results of the model account for experimental data on the observation of nanograins with 90° grain boundaries in YBaCuO films, reported in the literature.

Experimental results obtained for the atomic and magnetic distributions in the ferromagnetic Fe_0.865V_0.135 system are presented. These results were obtained from the observed polarized neutron scattering from a single crystal at room temperature. The results are compared with those obtained from a first-principles density functional based theory. This theory can calculate both the atomic short-range order that exists in atomically disordered alloys and also determine the effect that fluctuations in the local chemical environment have on the magnetization in ferromagnetic alloys. The present neutron scattering results for the atomic correlations in the ferromagnetic Fe_0.865V_0.135 system are in good agreement with those obtained from both a theoretical calculation and previous neutron scattering studies of FeV alloys. The current experimental results for the magnetic distribution in ferromagnetic Fe_0.865V_0.135 are seen to accurately reproduce the result determined from the theoretical calculation.

High-precision crystal structure parameters have been determined from Rietveld refinement of high-resolution neutron powder diffraction data in Nd₃(Fe,Ti)₂₉ and Nd₃(Fe,Ti)₂₉N_x intermetallics. The construction of the corresponding Wigner–Seitz (WS) polyhedra has revealed that some 12-fold (group-A) environments and the 14-fold (group-D, dumbbell) environments remain intact upon nitrogen uptake whereas some other 12-fold (group-B, C) environments transform according to the Frank–Kasper sequence . For comparison, WS cells were calculated for the 12-, 13-and 14-fold Frank–Kasper-like polyhedra of 1:12 and 2:17 types of structure. A global character of WS volume (WSV) relations has been revealed, where the corresponding WSVs follow the same trend for the three types of structure. Disclination nets have been constructed for the three types of structure, revealing that the easy-magnetization direction (EMD) of the iron sublattice is always along the [–Nd–Fe–] type of disclination line. A comparative analysis of the site hyperfine fields B_hfⁱ in the three kinds of compound led to successful estimations of saturation magnetization (M_s(Fe) = 42.6 μ_B/fu) from Fe-site magnetic moments for and its nitride (M_s(Fe) = 52.8 μ_B/fu) at room temperature. The successful prediction of Fe-site magnetic moments, of EMD and of M_s from the comparative study of Frank–Kasper-like environments shows that this approach can provide adequate guidance in the search for better rare-earth–transition-metal (TM = Fe,Co) permanent magnets.

The nature of spin freezing in geometrically frustrated icosahedral quasicrystals Tb–Mg–Zn and Tb–Mg–Cd was studied by thermoremanent dc magnetization (TRM) decay as a function of aging time and magnetic field. At low temperatures the magnetization exhibits typical broken-ergodicity phenomena, as characteristic of spin glasses (SGs). However, the observed linear dependence of the TRM on the magnetic field in the low-field regime is incompatible with the aging of a nonergodic system in an ultrametrically organized free energy of a SG, but compatible with a single-global-minimum free energy of a superparamagnet below the blocking temperature. The Tb–Mg–Zn(Cd) quasicrystals are, from this point of view, different from site-disordered SGs, but similar to geometrically frustrated pure (site-ordered) systems, like the kagomé and pyrochlore antiferromagnets, which also exhibit a superparamagnetic component in the magnetization below the spin freezing temperature and clustering of spins. The Tb–Mg–Zn(Cd) quasicrystals show features associated with both the site-disordered SGs and the superparamagnets. This duality is not a specific feature of spins in a quasiperiodic structure, but is found quite commonly in nonrandom (site-ordered) geometrically frustrated magnetic systems.

Magnetic hysteresis minor loops were measured with step by step increase of the magnetic field amplitude, H_a, in plastically deformed Fe single crystal. In order to analyse minor loops in connection with the lattice defects, we defined some magnetic parameters, such as the pseudo-coercive force, H_c^*, and the differential susceptibility under the pseudo-coercive force χ_H^*. H_c^*, for instance, is the magnetic field where the magnetization becomes zero in minor loops. In this work, we found the relationship 2H_c^* = H_a to hold over a fairly wide H_a range. These parameters are important for representing Bloch wall displacement and the potential energy. The parameters are remarkably sensitive to lattice defects in the field range below the coercive force H_c. The relation of 1/χ_H^* and H_c^* shows better sensitivity—90 times higher than that for H_c—for revealing information on dislocations. In the minor loop measurement, for getting full information on dislocations it is sufficient to have H_a = 400 A m⁻¹.

Polycrystalline Fe₃O₄ films have been prepared by reactive sputtering at room temperature. A transmission electron microscope image shows that the films consist of Fe₃O₄ grains well separated by grain boundaries with long-range and atomic scale disorders. The width of the long-range disordered grain boundaries observed by high resolution transmission electron microscopy is about , which is consistent with the boundary dimension derived from the resistivity and magnetization measurements. The temperature dependence of resistivity indicates that the transport properties of the films are dominated by the mechanism of fluctuation-induced tunnelling of electrons across the grain boundaries. Ordinary and extraordinary Hall constants were measured to be −8.22 × 10⁻¹¹ and , which are two and three orders of magnitude larger than those of iron, respectively.

The dependence of optically stimulated luminescence spectra on the energy of excitation photons has been investigated for a wide range of feldspars. In the vicinity of the 1.44 eV resonance, the spectra are in many cases best described by a Voigt profile with a minor Lorentzian component centred at 1.44 eV. Most samples display a second excitation resonance near 1.57 eV, and more rarely a third peak at 1.3 eV. The shape of the excitation spectrum is the same for both the violet (3.1 eV) and yellow–green (2.2 eV) emission bands and, for one sample, the UV band (3.3–3.7 eV). The shape of the excitation spectrum is unaffected by the polarization direction of the excitation beam. The principal resonance at 1.44 eV broadens little, if at all, with increasing temperature over the range 290–490 K. The initial luminescence decay rate as a function of the initial luminescence intensity per unit energy was determined over a wide range of excitation energies (1.24–2.58 eV) for four feldspars. In every case, the initial slope was observed to increase linearly with the square of the initial intensity over the range 1.24–2.4 eV. All this provides strong evidence that excitation is from a single trap located at least 2.5 eV below the conduction band.

A new high-sensitivity wide-bandwidth 1.25–5.5 eV (225–1000 nm) spectrometer has been constructed to measure luminescence emission spectra of minerals that are of interest for optical dating. Spectra of emission resulting from 1.43 eV (IR) excitation after γ-irradiation are reported for 13 cut rock feldspars and 20 feldspar separates. Also reported are phosphorescence spectra following γ-irradiation, and after 1.4 eV excitation. The main differences between the infrared stimulated luminescence spectra and the phosphorescence is the almost complete absence of the violet, 3.1 eV, and yellow–green, 2.2 eV, bands in the phosphorescence, and the presence of a green emission band centred at 2.7 eV in the phosphorescence following γ-irradiation (but absent in the phosphorescence following 1.4 eV excitation). The red, 1.7 eV, band is present in all the phosphorescence spectra but is not always seen during 1.4 eV excitation. A band at  eV is dominant in both types of phosphorescence spectra. This dependence of the luminescence spectrum on the mode of excitation suggests a strong correlation between certain traps and luminescence centres. Models involving recombination via a 'conduction band' in the traditional sense cannot account for these observations.

Under a static high magnetic field of 10 T, diamond-like carbon (DLC) nanocrystals and graphite-coated n-diamond nanoparticles have been synthesized after a pyrogenation of carbon black and a nanometre-sized iron catalyst at atmospheric pressure and a temperature of 1100 °C. The product is analysed by x-ray diffraction, Raman spectroscopy, transmission electron microscopy and electron-probe microanalysis. The average size of the DLC nanopowders is about 20 nm, and that of the graphite-coated n-diamond particles is about 100 nm. The yield of diamond is as high as 30%.

We propose a model for non-ideal monitoring of the state of a coupled quantum dot qubit by a quantum tunnelling device. The non-ideality is modelled using an equivalent measurement circuit. This allows realistically available measurement results to be related to the state of the quantum system (qubit). We present a quantum trajectory that describes the stochastic evolution of the qubit state conditioned by tunnelling events (i.e. current) through the device. We calculate and compare the noise power spectra of the current in an ideal and a non-ideal measurement. The results show that when the two qubit dots are strongly coupled the non-ideal measurement cannot detect the qubit state precisely. The limitation of the ideal model for describing a realistic system may be estimated from the noise spectra.