Table of contents

Considering the magnetic ordering observed in Fe_1−xCo_xSi compounds, the inhomogeneity effect on the Curie temperature of the inhomogeneous system is studied by using the coherent potential approximation. In the model system introduced in the present study, the magnetically ordered state can be realized in a finite concentration range between the insulating spin gap state corresponding to FeSi and the metallic paramagnetic state corresponding to CoSi. The concentration dependences of the density of states and Curie temperature are the main focuses of investigation. The density of states at the Fermi level is strongly suppressed by the inhomogeneity effect. The suppression of the density of states also leads to suppression of the transition temperature. Although the electron correlation effect generally suppresses the Curie temperature similarly, the inhomogeneity effect on the suppression of the transition temperature cannot be replaced with the correlation effect of the effective Coulomb interactions.

A brief review is given of optical studies of doped II–VI quantum wells. The properties of exciton states, neutral as well as positively and negatively charged, are discussed. A wide range of optical measurements is presented: CW as well as picosecond and femtosecond time-resolved absorption, photoluminescence (PL) and PL excitation. The experiments were performed at various carrier concentrations (> 10¹¹ cm⁻²) and temperatures (up to a few tens of kelvins). This review is limited to zero or low magnetic fields, used only to polarize spins of carriers. We discuss the obtained values of various fundamental parameters of the excitonic states: energies, optical transition probabilities and characteristic times of their formation, thermalization and decay.

This article reviews the current status of chemically doped fullerene superconductors and related compounds, with particular focus on Mott–Hubbard states and the role of molecular orbital degeneracy. Alkaline-earth metal fullerides produce superconductors of several kinds, all of which have states with higher valence than (C₆₀)⁶⁻, where the second lowest unoccupied molecular orbital (the LUMO + 1 state) is filled. Alkali-metal-doped fullerides, on the other hand, afford superconductors only at the stoichiometry A₃C₆₀ (A denotes alkali metal) and in basically fcc structures. The metallicity and superconductivity of A₃C₆₀ compounds are destroyed either by reduction of the crystal symmetry or by change in the valence of C₆₀. This difference is attributed to the narrower bandwidth in the A₃C₆₀ system, causing electronic instability in Jahn–Teller insulators and Mott–Hubbard insulators. The latter metal–insulator transition is driven by intercalation of ammonia molecules into A₃C₆₀-type superconductors. Furthermore, the triple degeneracy of the LUMO state of C₆₀ plays a crucial role in the metal–insulator transition and in controlling the magnetic structures of insulating states, possibly providing novel properties of degenerate orbitals. The goal of this article is to establish a unifying picture of fullerene intercalation compounds, and to clarify the underlying physics: competing energy scales and orbital properties of molecule-based systems.

Interplay of nonlinear and quantum effects in the ground state of the E ⊗ (b₁ + b₂) Jahn–Teller model was investigated by the variational approach and exact numerical simulations. They result in the recognition of (i) the importance of the admixture of the first excited state of the displaced harmonic oscillator of the symmetric phonon mode in the ground state of the system in the self-trapping-dominated regime, and (ii) the existence of the region of localizedb₁-undisplaced oscillator states in the tunnelling-dominated regime. Effect (i) occurs owing to a significant decrease in the ground state energy on account of the overlapping contribution of the symmetric phonon mode between states of the same parity. This contribution considerably improves variational results, especially in the self-trapping-dominated regime. Close to the E ⊗ e limit, the nonlinear effects of two-mode correlations turn out to be effective due to the rotational symmetry of this case. In the tunnelling-dominated regime the phonon wavefunctions behave like the strongly localized harmonic oscillator ground state and (i) loses its significance.

This paper presents an experimental measurement of the electronic structure of Na₂O in the solid phase using electron momentum spectroscopy and compares the results with ab initio calculations performed within the linear combination of atomic orbitals (LCAO) approximation. While Hartree–Fock (HF) can reproduce elastic properties we find it overestimates splitting of the oxygen valence bands by around 30% and the width of the O 2p band by a factor of 2. Our experimental values are 15.85 ± 0.2 and 0.6 ± 0.2 eV for these two quantities, respectively. Density functional methods are significantly better, with the hybrid functional PBE0 predicting the oxygen bandgap to within the experimental error. PBE0 also gives the best estimate of the Na core level energies. In contrast, HF performs best for the splitting between the oxygen and sodium bands. Our experimental values of 32.85 ± 0.2 and 27.45 ± 0.2 eV for the Na 2p–Na 2s and O 2p–Na 2p splittings agree well with previous measurements. Distribution of electron density both within the bands and between the bands is not reproduced by any of the computational methods employed.

Low temperature (7 K) absorption spectra of polycrystalline sintered ceramic Lu₂O₃:Eu³⁺ were recorded between 3950 and 50000 cm⁻¹. There are two different intrinsic Eu³⁺ sites, C₂ and S₆ (C_3i), in this host. A total of 105 crystal-field (CF) energy levels were assigned and fitted to a semiempirical Hamiltonian representing the combined free-ion (FI) and CF interactions for a 4f⁶ ion in the C₂ symmetry site. 10 FI and 14 CF parameters were varied simultaneously in the least square adjustments yielding an rms deviation between the calculated and experimental levels of 15 cm⁻¹. The CF strength parameter, S, obtained from calculated B_q^k parameters is larger for Lu₂O₃ when compared to the Y₂O₃ host, which is in accordance with the smaller ionic radius of the Lu³⁺ ion. The CF splittings of the ⁷F ₁ and ⁵D ₁ levels were also determined experimentally for the Eu³⁺ ion in the centrosymmetric S₆ site and the value of the second-rank B₀² CF parameter was calculated.

This work investigates the photoluminescence (PL) properties of the cubic chloroperovskite NH₄MnCl₃. Like in most concentrated materials, the Mn²⁺ PL which is located at 2.10 eV at T = 10 K strongly depends on the temperature. Optical absorption (OA), emission, and excitation spectroscopy, as well as lifetime measurements, performed on NH₄MnCl₃ indicate that the PL is mainly intrinsic at T = 10 K and consists of a broad band located at 2.10 eV. Above this temperature, the PL gradually transforms to extrinsic PL due to exciton migration and subsequent trapping. Further temperature increase above 100 K yields transfer to killers of excitation which are responsible for the PL quenching, and hence the absence of PL at ambient conditions. The exciton traps are identified with perturbed Mn²⁺ sites with the effective activation energy of 52 meV, whilst the activation energy for energy transfer is 47 meV. The existence of these traps has been directly revealed by time-resolved spectroscopy. The detected intrinsic and extrinsic PL bands are displaced by 6 meV, which coincides with the activation energy difference between pure Mn²⁺ and trap Mn²⁺, as derived from temperature dependence studies of the lifetime τ(T). Interestingly, a PL band at 1.82 eV is observed above 60 K. This band, which was initially associated with deeper excitation traps, actually corresponds to precipitates of MnCl₂ inside NH₄MnCl₃. The correlation analysis performed on NH₄MnCl₃ using OA, PL, and lifetime data provides an estimate of the precipitate concentration of 0.3 mol%. The presence of two separated Mn²⁺ PL bands at different temperatures is a rather common phenomenon in concentrated materials such as AMnX₃ (A = NH₄, Rb; X = Cl, F), and has been interpreted in terms of exciton transfer to deeper traps. The present finding stresses the relevance of an adequate structural characterization in dealing with PL in concentrated materials.

Electrical conduction at high fields was examined in a series of hydrogenated amorphous silicon oxynitride and silicon nitride films with different nitrogen contents deposited by plasma-enhanced chemical vapour deposition. It was shown that the conduction is attributable to the Poole–Frenkel (PF) emission in the two materials. The energy depths of the PF sites and the dependences on the sample's chemical composition are quite similar for the two samples. It is considered that the PF sites in the two materials are identical.

We present a Monte Carlo study of the helicity moduli of an anisotropic classical three-dimensional XY-model of YBCO in the superconducting state. It is found that both the ab-plane and the c-axis helicity moduli, which are proportional to the inverse square of the corresponding magnetic field penetration depth, vary linearly with temperature at low temperatures. The result for the c-axis helicity modulus is in disagreement with the experiments on high-quality samples of YBCO. Thus we conclude that purely classical phase fluctuations of the superconducting order parameter cannot account for the observed c-axis electrodynamics of YBCO.

The crystal structure, Curie temperature and magnetostriction of compounds with the formula Nd_0.67Tb_0.33 (Co_1−xFe_x)₂ were investigated by x-ray powder diffraction and magnetic measurements. It is revealed that the substitution of Fe for Co destabilizes the cubic Laves phase Nd_0.67Tb_0.33 (Co,Fe)₂. The maximum substitution of Fe for Co is about 0.57. The lattice constant of the Nd_0.67Tb_0.33 (Co,Fe)₂ phase increases with x increasing from x = 0.0 to 0.6, then remains almost unchanged from x = 0.6 to 0.9. The Curie temperature exhibits a variation similar to that of the lattice constant with the Fe concentration. A spin reorientation is observed, which can be understood on the basis of the two-sublattice model. The spontaneous magnetostriction constant λ₁₁₁ shows a maximum value of 1.7 × 10⁻³ at x = 0.4. The temperature dependence of λ₁₁₁ for Nd_0.67Tb_0.33 (Co_0.6Fe_0.4)₂ exhibits an abrupt jump at 165 K, which corresponds to the temperature of the spin reorientation.

Large-scale zinc oxide (ZnO) nanowires were grown via a simple chemical reaction involving water vapour. Electron microscopy observations reveal that the ZnO nanowires are single crystalline and grow along the c-axis ([001]) direction. Room temperature photoluminescence measurements show a striking blue emission at 466 nm along with two other emissions in the ultraviolet and yellow regions. Annealing treatment of the as-grown ZnO nanowires results in an apparent reduction of the intensity of the blue emission, which indicates that the blue emission might be originating from the oxygen or zinc defects generated in the process of growth of the ZnO nanowires.

In a recent paper, Zenkov and Moskvin (2002 J. Phys.: Condens. Matter14 6957) analysed the influence of bismuth on magneto-optical effects in iron garnets, questioning the validity of previous approaches (Dionne and Allen 1993 J. Appl. Phys.73 6127; 1994 J. Appl. Phys.75 6372, Allen and Dionne 1993 J. Appl. Phys.73 1630, Helseth et al2001 Phys. Rev.64 174406). In this comment I point out that these claims apparently have no foundation.

The discussion concerns the origin of the giant Faraday rotation in bismuth- and lead-doped iron garnets. It is convincingly shown that this effect is due to the covalent admixture of Bi(Pb)6p-wavefunctions to oxygen 2p-orbitals in octahedral and tetrahedral Fe–O clusters of iron garnets. The crucial role of the quantum-chemical computation of electronic structure of such clusters is emphasized.

In figure 2 of our paper, the uppermost HREEL spectrum, relative to oxygen adsorption on Ag(410), was recorded at a negative, instead of positive, angle of incidence as stated in the paper. It corresponds therefore to the adsorption of molecules hitting the surface close to the normal to the (100) nanofacets and not, as stated erroneously in the caption, to the (110) nanofacets. This error has no further consequences for the discussion and the conclusions, since on Ag(210) oxygen adsorption is dissociative for all angles of incidence.

The authors regret that figure 7 on page 3153 of this article was mislabelled. The numbers labelling the peaks were plane-wave frequencies (ω_kβ), not energies (ε_kβ=ω²_kβ) as indicated. A corrected version of the figure is given below. The conclusions are unaffected.

The spectral densities for vector vibrations in the fcc lattice with force-constant disorder (the width of the box distributions Δ = 1) calculated by the CPA (solid curves) and KPM (dashed curves).