Table of contents

High-resolution powder neutron diffraction has been used to study the crystal structure of the Mg¹¹B₂ superconductor (T_c = 39.2 K) prepared at high pressure in the temperature range 2 to 293 K. The experimental data provide clear evidence of temperature-independent phase inhomogeneity, which was modelled by coexisting Mg_1+δB₂ phases. Precise structural information for temperatures near T_c also reveals extremely small volume and c-axis discontinuities, consistent with the existence of a specific heat anomaly at T_c.

Zr₇₀Cu₃₀ and Zr₇₀Cu₂₀Rh₁₀ glassy alloys were prepared by a single-roller melt-spinning method and their crystallization processes were studied by differential scanning calorimetry, x-ray diffraction and transmission electron microscopy. For the Zr₇₀Cu₃₀ alloy, the glassy state crystallizes through a single exothermic reaction due to the precipitation of the stable Zr₂Cu crystalline phase. For the Zr₇₀Cu₂₀Rh₁₀ alloy, metastable icosahedral quasicrystalline and FCC-Zr₂Rh phases precipitate from the glassy matrix in the initial crystallization process, followed by the phase transformation to Zr₂Cu and ZrRh crystalline phases upon further annealing. These results illustrate that the partial substitution of Cu by Rh in the Zr₇₀Cu₃₀ glassy alloy is effective in promoting the precipitation of the metastable icosahedral quasicrystalline phase. To explain these experimental results, the existence of icosahedron-like atomic clusters in the Zr₇₀Cu₂₀Rh₁₀ glassy alloy is suggested.

The isothermal magnetization m(H) of the metamagnet FeCl₂ is measured in axial magnetic fields 0⩽µ₀H_a⩽12 T at temperatures 34⩽T⩽53 K above the Néel temperature, where the system is essentially a two-dimensional Ising ferromagnet. The analysis of the data indicates experimental accessibility of the critical exponent µ of the Yang-Lee edge singularities. They manifest themselves in divergences of the density functions g(θ), which quantify the distribution of the zeros of the partition function on the Lee-Yang unit circle in the complex plane. In accordance with the hypothesis of universality, a critical exponent close to the theoretical prediction for the two-dimensional Ising ferromagnet, µ = -1/6, is found.

A general concept for photoinduced structural phase transitions is developed, in terms of the hidden multi-stability of the ground state and the proliferations of optically excited states. Taking the ionic (I) to neutral (N) phase transition in the organic charge transfer (CT) crystal, tetrathiafuluvalene-p-chloranil, as a typical example for this type of transition, we, at first theoretically show an adiabatic path which starts from CT excitons in the I-phase, but finally reaches to a N-domain with a macroscopic size. In connection with this I-N transition, the concept of the initial condition sensitivity is also developed so as to clarify experimentally observed nonlinear characteristics of this material.

Then, using a simplified model for the many-exciton system, we theoretically study the early-time quantum dynamics of the exciton proliferation, which finally results in the formation of a domain with a large number of excitons. For this purpose, we derive a stepwise iterative equation to describe the exciton proliferation, and clarify the origin of the initial condition sensitivity.

Possible differences between a photoinduced non-equilibrium phase and an equilibrium phase at high temperatures are also clarified from general and conceptional points of view, in connection with recent experiments on the photoinduced phase transition in an organo-metallic complex crystal. It will be shown that the photoinduced phase can make a new interaction appear as a broken symmetry only in this phase, even when this interaction is almost completely hidden in all the equilibrium phases, such as the ground state and other high-temperature phases.

We discuss non-Fermi liquid and quantum critical behaviour in heavy-fermion materials, focusing on the mechanism by which the electron mass appears to diverge at the quantum critical point. We ask whether the basic mechanism for the transformation involves electron diffraction off a quantum critical spin-density wave, or whether a breakdown in the composite nature of the heavy electron takes place at the quantum critical point. We show that the Hall constant changes continuously in the first scenario, but may `jump' discontinuously at a quantum critical point where the composite character of the electron quasiparticles changes.

Equilibrium segregation of sulfur to the free surface of single crystalline titanium from 560{°}C to 800{°}C was investigated using Auger electron spectroscopy (AES) measurements. To describe the concentration evolution in the sulfur overlayer, Fick's first law was modified by adding a local function df(x)/dx, to the concentration gradient to drive the segregation starting from an initially homogeneous distribution. The diffusion equation thus derived was solved for the case f(x) = a_sexp (-x/d_s). It was found that the solution leads to an AES intensity evolution for segregants, I_S(t) = I_S^∞(1-e^αerfc((α)^1/2)), α = Dt/d_s², which fits the experimental results extremely well. An exponentially decaying distribution of sulfur beneath the titanium surface at equilibrium was revealed by sputter depth profiling, which in turn justifies our modification to Fick's first law. Without referring to the detailed kinetics, an activation energy E_a = 718 meV/atom was determined by comparing the sulfur concentration evolution at different temperatures.

A theoretical model which describes the generation of misfit dislocations in multilayered films deposited onto plastically deformed substrates with disclinations (defects of rotational type) is suggested. In the framework of the model, the ranges of the parameters (disclination strength, misfit parameters, layer thicknesses, characteristics of disclination arrangement) of a multilayered film/substrate composite for which the generation of misfit dislocations is energetically favourable are calculated. The specific features of the generation of misfit dislocations in multilayered films on disclinated substrates are discussed in relation to a technologically interesting possibility for exploiting plastically deformed substrates.

We report our theoretical considerations motivated by Brillouin spectra of polymethyl methacrylate (PMMA) films (d = 13 nm and d = 225 nm) on Si(100) substrates. We have rederived a numerical calculational scheme developed for considering Brillouin experiments on layer systems. It delivers dispersion curves and intensities, indicated here by appropriate hatchings. With backscattering geometry, we observed guided waves in agreement with our calculations. The Brillouin peaks fall on the calculated segments of the dispersion curves visible with the elasto-optic mechanism for 180° scattering or for reflection-induced `A-scattering' (called RIΘA scattering by Krüger et al (Krüger J K, Embs J, Brierley J and Jimenez R 1998 J. Phys. D: Appl. Phys. 31 1913)). The observed pseudo-guided waves for d = 225 nm were explained using a model of a PMMA film fixed at one surface and free at the other. For the d = 225 nm thick film the Brillouin data provided information on its mechanical constants, but in the d = 13 nm case the dispersion curves are not very sensitive to the film properties. The Brillouin data for d = 13 nm show the direction of the sagittal plane with respect to the symmetry of the silicon rather than indicating the properties of the PMMA film.

The melting behaviour of Pb film embedded in an Al matrix has been investigated by a molecular dynamics simulation technique. The Sutton-Chen potential for alloys has been employed to perform these calculations. Melting of a Pb(110) film has been studied by embedding it in Al(110) and Al(111). Melting is found to initiate at the film-host interface for the Al(110) host. However, a change in orientation of the film is observed before superheating when the film is embedded in the Al(111) matrix. The superheating determined by simulation is comparable with the value calculated by homogeneous theory.

The electron-doped infinite-layer superconductor Sr_0.9La_0.1CuO₂ is studied with x-ray photoemission spectroscopy (XPS). A non-aqueous chemical etchant is shown to effectively remove contaminants and to yield surfaces from which signals intrinsic to the superconductor dominate. These data are compared to measurements from hole-doped La_1.85Sr_0.15CuO₄, from undoped La₂CuO₄, and from electron-doped Nd_1.85Ce_0.15CuO_4-δ. The Cu 2p core level is more consistent with a lower value of the O 2p→Cu 3d charge transfer energy Δ than in hole-doped cuprates. A clear Fermi edge is observed in the valence band region.

Recent experiments on magnetic tunnelling devices based on magnetite film show an unexpectedly low tunnel magnetoresistance despite a high magnetic transition temperature and a high spin polarization of the conduction electrons. The explanation is given in two parts. The first reason for the reduction is the quantum mechanical coupling and relaxation of the conduction electron and core spin. This can reduce the tunnel magnetoresistance substantially. The second is the surface effects in which the magnetic spins at the surface are frustrated as a result of fewer nearest neighbours, surface reconstructions, anti-phase boundaries and the superparamagnetic behaviour. This explains the rapid fall of the magnetization and the tunnel magnetoresistance with temperature.

The energy transfer efficiency from Er³⁺ to Yb³⁺ ions in yttrium orthovanadate single crystals (YVO₄) is experimentally obtained, by using a method based on the simultaneous and multiwavelength measurement of photoacoustic and luminescent signals after pulsed laser excitation. The result is reached by comparing with the predictions from Judd-Ofelt analysis and the lifetime measurements. The energy transfer between the ions, from Er³⁺ to Yb³⁺, must be considered in order to fit the experimental results. A value of energy transfer efficiency (Ψ = 0.16) is obtained.

A new sample of HoPd₂Si₂ (ThCr₂Si₂-type structure, space group I4/mmm) has been prepared and investigated. Magnetic susceptibility and magnetization data indicate antiferromagnetic behaviour below T_N = 5.8 K. On the basis of powder neutron diffraction data the magnetic structure of the compound has been redefined and found to be more complex than reported earlier. In the temperature region from 20 K to 5 K a short-range magnetic order develops. Between 1.5 K and 5 K a long-range magnetic ordering, that could be described in an enlarged magnetic unit cell (a_m = 7a, b_m = a and c_m = 6c), is observed. A complex magnetic structure of sine-wave modulated type with two propagation vectors and Ho magnetic moments parallel to the b-axis is proposed. Moreover, a new attempt to understand the transition from long- to short-range magnetic ordering observed in some rare earth compounds with Pd is made.

The local lattice distortions around a Tl⁺-dimer substitutional impurity in NaI and KI have been investigated by using a mixed ab initio/parametrized methodology. One important conclusion of the work is that an explicit consideration of these distortions up to at least the first four coordination shells of ions around the impurity is needed in order to achieve a converged result for the first- and second-shell distortions. After describing the lattice distortions induced by both D_4h and D_2h configurations, we analyse their relative stability by calculating the defect formation energies. It is found that the D_4h configuration is lower in energy for both crystals, but the energy difference is so small in the case of the NaI host lattice that the two configurations can be considered degenerate for all practical purposes. It is also found that inclusion of polarization interactions is crucial in the stabilization of the defect in NaI.

This paper reports a new wave phenomenon. We show that in a system consisting of many electric dipoles, the interaction between dipoles, mediated by radiated electromagnetic waves, leads to a kind of phase transition in the system. When the phase transition occurs, all dipoles tend to oscillate in phase, displaying a global coherent behaviour. It is also found that this feature is independent of the precise configuration of the dipole system.

This work applies a recently developed first-principles scheme for calculating core excited states in solids to non-resonant inelastic x-ray scattering. The model explicitly includes the interaction between the excited electron and the core hole. The calculated results are compared with recent experimental results on inelastic x-ray scattering from K shells of Li, Be and C in lithium fluoride, beryllium oxide and diamond, respectively. The overall agreement between experimental and theoretical spectra is found to be good in all cases, even for the near-edge structure.

Using a generalized tight-binding model, we study the changes induced in the electronic and mechanical properties of carbon nanotubes with encapsulated C₆₀s. Provided enough overlap exists between the electronic states of the nanotube and those of the C₆₀s, a tiny gap (~0.01-0.02 eV) opens in the band structure of a metallic tube. The gap is seen to be wider for smaller separations between the C₆₀s. For semiconductor tubes, on the other hand, the encapsulated C₆₀s produce donor levels in the gap causing it to narrow. As regards mechanical properties, doped tubes are observed to be slightly `softer' than undoped ones. This is indicated by reductions of the Young modulus and torsional rigidity of the doped tubes by 0.4-1.8% and 0.6-1.2%, respectively, as compared to those of the pure tubes. Moreover, the Poisson ratio of the doped tubes is seen to be lower by ~5%. These novel features of the fullerene-doped nanotubes should be of interest in future applications.

Coalescence of three identical size nanoclusters, each containing 309 silver atoms interacting through an analytic embedded-atom method (EAM) type potential, is studied by molecular dynamics simulations at various temperatures. The rotation and spontaneous self-organization of all three clusters in the coalescence process are observed. The final low-energy configurations of the coalescing system at low temperatures (T⩽800 K) are constituted of three particles separated by attached interfaces where edge dislocations are formed, whereas the coalescing at high temperatures (T>800 K) gives rise to a single-like particle. The melting temperature of the coalescing system is well below that of the free single cluster with the same particle size.

The energies of 135 Kramers doublets extending up to the ²H1_11/2 multiplet for Nd³⁺ in a monoclinic C2/c space group (No 15) NdAl₃(BO₃)₄ (NAB) single crystal laser have been determined from polarized optical absorption and photoluminescence measurements at 7 K. The strongly polarized character of the Nd spectra has been discussed under the assumption of a local D₃ symmetry, higher than the C₂ symmetry of NAB, and the observed energy levels have been labelled with the adequate crystal quantum numbers and irreducible representations. A detailed Hamiltonian of 21 parameters has been used in the simulation of the energy levels and associated wavefunctions of the 4f³ configuration of Nd³⁺. The diagonalized complete energy matrix combines simultaneously the free-ion and single-particle crystal field interactions. Starting B^k_q CF parameters were calculated from the semi-empirical simple overlap model SOM. A comparative simulation considering the C₂ symmetry of NAB is provided. Moreover, two-electron CF interactions as well as an empirical correction have been tested in calculating the anomalous splitting of the ²H2_11/2 levels. A final fit in D₃ symmetry produces a very good adjustment with a low rms deviation σ = 15.3 cm^-1 between observed and calculated energy levels.

This paper discusses a model for dissipation in electrical conductors, in which a time-dependent flux X(t) is threaded through a loop. The rate of dissipation is proportional to a constant D_E characterizing the diffusion of single-particle energies, and Ohmic dissipation occurs when D_E∝². The diffusion constant D_E is investigated for a random-matrix model, and compared with the predictions of the Kubo-Greenwood formula. The results agree with the Kubo-Greenwood formula in a physically important regime where other random-matrix models do not show agreement. Numerical experiments also identified another regime of this model which exhibits Ohmic dissipation, with a conductance which does not agree with the Kubo-Greenwood formula.

In the present work we study the defect states in recently synthesized semiconductor quantum dot–quantum well systems (QDQWs). We employ the effective-mass theory (EMT) with a realistic barrier and variable effective mass. The model is simple and all of our results are obtained by an exact numerical diagonalization of the Schrödinger equation. We study the ground state of the host system as a function of quantum well size. We demonstrate that the upshift with QDQW downsizing differs from reported upshifts in simple quantum dots (QDs) and explain this by means of a perturbative analysis. We study the binding energy of the hydrogenic impurity and its variation with QDQW size. Next the binding energy is studied as the impurity is moved off-centre. We find that the binding energy goes through a maximum. An analysis of the wavefunction is carried out to obtain an understanding of this surprising effect. The impurity calculations are carried out on CdS/HgS/CdS QDQWs. Recently, experimental studies on a monolayer of HgS in a CdS dot were carried out. We model this system as a `δ-defect' consisting of a thin spherical shell of fixed potential depth in a spherical CdS dot. The ground state of this system is studied as the shell is dragged from the periphery to the centre of the CdS dot. Our results are explained on the basis of qualitative arguments and asymptotic analyses. Outstanding issues and future directions are suggested.

Persistent currents and Drude weights are investigated using the tight-binding approximation to one-dimensional rings threaded by a magnetic flux and with the potential given by some almost-periodic substitution sequences with different degrees of randomness, and for various potential strengths. The Drude weight D distinguishes correctly conductors and insulators, in accordance with the results indicated by the currents. In the case of insulators the decay of D(N) for large ring lengths N provides an estimate for the localization length of the system. It is shown that a more random sequence does not imply reduced conduction. This discrepancy between the hierarchy of disorder of the sequences and the capacity for conduction of the system is explained by the gaps in the energy spectra.

The magnetoresistance (MR) of amorphous indium oxide films on the insulating side near the superconductor-insulator transition was measured. Variable range hopping is evident in the presence of a high enough magnetic field even in the temperature range where simple activation prevails in the absence of a magnetic field, which strongly suggests the existence of localized superconducting granules. Consequently, junction breaking between superconducting granules and pair breaking effects dominate the MR at low enough temperatures. As those effects caused by the local superconductivity on MR decrease rapidly with increasing temperature, an intrastate interaction effect becomes significant. The observed MR is fitted to a theoretical expression which includes junction breaking, pair breaking and intrastate interaction terms. The temperature dependence of the fitting parameters shows qualitative agreement with theoretical expectations.

Polarization dependent electrolyte-electroreflectance (EER) measurements of ReS₂ and ReSe₂ layered crystals have been carried out in the energy range of 1.3 to 5.0 eV. The EER spectra of E∥b polarization exhibit distinct band-edge excitonic and interband transition features from those of E⊥b polarization. Analysing the polarization dependent EER spectra, the structures of the excitonic and interband transitions of ReS₂ and ReSe₂ with optical polarizations along the b-axis and perpendicular to the b-axis are examined clearly and the energy positions are determined accurately. The mechanism of field-lattice interaction is proposed to account for the in-plane anisotropy of the crystals. Based on the analysis of experimental results, a probable band-structure scheme of ReX₂ (X = S, Se) is constructed.

We study the magnetocrystalline anisotropy of Fe_0.5Pd_0.5 alloy using the first-principles spin-polarized relativistic Korringa-Kohn-Rostoker coherent-potential approximation. We investigate the effect of long-ranged chemical order on the magnitude as well as the direction of easy magnetization. We find that in this alloy, the chemical order enhances the magnitude of the magnetocrystalline anisotropy energy as well as altering the easy axis. In particular, for the L1₀ ordered alloy the easy axis is perpendicular to the alternating layers of Fe and Pd with quite large magnetocrystalline anisotropy energy 124 µeV. These observations are in very good agreement with experiments. We also find that it is the electronic structure in the vicinity of the Fermi surface that is responsible for this effect.

The magnetic properties and ⁵⁷Fe Mössbauer spectra of the compounds YFe₁₁TiH_x (x = 0, 1) were investigated. The magnetocrystalline anisotropy of YFe₁₁Ti and its hydride was studied by analysing the hard and easy magnetization curves of single crystal samples in the temperature range 4.2–300 K. The spontaneous magnetization was determined in wide temperature range 4.2–650 K. It is established that at T = 4.2 K magnetic anisotropy constants K₁ and K₂ for YFe₁₁TiH₁ single crystal reach values of 25.8 K f.u.⁻¹ and 0.24 K f.u.⁻¹, respectively. The Mössbauer effect spectra of YFe₁₁Ti and its hydride were analysed in terms of a model which takes into account the local environment of Fe atoms on three crystallographic sites (8f, 8j and 8i) and an influence of the random distribution of titanium on the 8i site. Upon hydrogenation both the hyperfine fields and the isomer shifts increase. These results are discussed in terms of the hydrogen-induced unit cell expansion and the electron charge transfer from the conduction band onto the H atoms.

CaFe_xTi_1-xO_3-y (0.05⩽x⩽0.60) and SrFe_xTi_1-xO_3-y (0.20⩽x⩽0.60) perovskites were prepared by solid state reaction and subjected to different heat treatments in different atmospheres. Mössbauer spectroscopy of the CaFe_xTi_1-xO_3-y could detect Fe⁴⁺ in all the samples, even in those heated at 1000 °C in Ar atmosphere for 10 hours. Fe³⁺ coordinated by six, five and four anions were also identified and the estimated hyperfine parameters were consistent with those previously reported for the CaFe_xTi_1-xO_3-x/2 oxides where all the Fe cations were present as Fe³⁺. The Fe⁴⁺/Fe³⁺ ratios estimated from Mössbauer data are in agreement with coulometric titration data. The absence of tetracoordinated Fe³⁺ in SrFe_xTi_1-xO_3-y suggests that in contrast to CaFe_xTi_1-xO_3-y no ordering of anion vacancies takes place even for x = 0.60 and explains the differences in the Fe concentration dependence of the electric transport properties of these materials.

We report our studies on the electronic structure and linear and nonlinear optical (NLO) properties of KNbO₃ using a first-principles method in the local density approximation (LDA). The calculated results for the refractive indices and second-harmonic-generation (SHG) coefficients agree well with experimental results. From decomposing the nonlinear susceptibility, we find that the primary contribution to the NLO behaviour comes from the hybridization of the O 2p and Nb 4d electron states. In addition, there are two different roles played by the O atoms because of their different distances from the Nb atom; thus we propose a possible way to enhance the SHG coefficients.

Excitonic effects on the third-harmonic-generation coefficient for disclike parabolic quantum dots are studied, and an analytic formula for the third-harmonic-generation coefficient is obtained by using the compact-density-matrix approach and an iterative method. Numerical results are presented for typical GaAs/AlGaAs parabolic quantum dots. The results show that the third-harmonic-generation coefficient is greatly enhanced because the excitons are localized in the quantum dots - it is over three times that obtained by just considering electron states - and that it is very sensitively dependent on the exciton confinement. In addition, the third-harmonic-generation coefficient is related to the relaxation constant.

Luminescence spectral and kinetic characteristics of CsPbCl₃ nanocrystals, dispersed in perovskite-like CsSrCl₃ matrix, have been studied under excitation by synchrotron radiation in the UV–visible spectral region (4 to 20 eV). Various time components in the decay kinetics of CsPbCl₃ nanocrystals appear due to direct excitation of nanocrystals or reabsorption of the luminescence of other emitting centres. Shortening of the luminescence decay time of the CsPbCl₃ nanocrystals with respect to that observed for a CsPbCl₃ single crystal has been explained as evidence of quantum-size effect.

To calculate the effective dielectric response of a dilute composite, a generalization of the Maxwell Garnett theory for small nonspherical particles distributed in shape is proposed. Various types of distribution function are analysed and the applicability of the simplest (steplike) distribution is discussed. It is shown that the use of the steplike distribution is more valid for particles having a higher imaginary part of the permittivity in the actual region. Besides, an alternative approach to the problem based on the spectral representation is also considered. As an illustration, the effective dielectric response of a system of semiconductor (SiC) and metal (Al) ellipsoidal particles is calculated.