Table of contents

A new mechanism for nucleation of dislocation dipoles at nanoparticles (nanoinclusions) in nanocomposite solids is suggested and theoretically described. The mechanism represents the nucleation of a nanoscale dipole of 'non-crystallographic' partial dislocations whose Burgers vector magnitudes continuously grow during the nucleation process. It is shown that the dislocations nucleated at nanoparticles can be emitted into a matrix in nanocomposites deformed at high mechanical stresses.

The structure and phase transition in lead zinc niobate (PZN) were studied using very high resolution powder neutron diffraction between 4.2 and 450 K. The structure is unequivocally rhombohedral in space group R3m with a = 4.060 48 Å and α = 89.8693° at 4.2 K. No low symmetry phases were observed, nor was an octahedral tilting transition to R3c as occurs in some parallel systems (e.g. PZT). Within the rhombohedral structure, large average ion displacements were found relative to the rather small spontaneous strains. On heating, a transition to a cubic perovskite occurs, with a critical temperature of ∼384 K.

Deposition of In onto a Bi nanoline template on the Si(001) surface at a very low flux produces long, flat islands with a zigzag appearance. We show that these islands result from the intermixing of the In and the Bi dimer atoms in the nanoline, and comprise an In–Bi atomic chain structure along each half of the nanoline in which the In and Bi atoms can take up favoured hybridizations—sp² (with empty p orbital) and p³ (with filled s orbital) respectively. There are six different possible isomers of the double-zigzag structure, resulting from different arrangements of the chains on the two sides of the nanolines. Tight-binding and density functional theory (DFT) simulations of the different candidates find that three of these structures have a higher energy than the other three as a result of a repulsive Bi–Bi Coulomb interaction. The two lowest-energy isomers will both match the appearance of the double zigzag in filled-state STM images.

We propose that the onset of a glass transition can be defined as the point at which a supercooled liquid acquires the stress relaxation mechanism of a solid glass. We translate this condition into the rate equation for local relaxation events. This equation simultaneously gives two main signatures of a glass transition: stretched-exponential relaxation and the Vogel–Fulcher law. In the proposed picture, the increase of non-exponentiality of relaxation and deviation from the Arrhenius law both originate from the lag behind the new equilibrium state, to which the system is driven by external perturbation. This lag can be quantified by the number of local relaxation events, and we propose that this number is related to the system's fragility. The proposed theory reproduces several general correlations, including the correlation of fragility with non-exponentiality of relaxation and bonding type.

We have measured the valence-band densities of states (DOSs) in buried CoFeB and CoFe layers in a magnetic tunnel junction using a novel extension of a recently developed standing wave/wedge soft x-ray photoemission technique. The CoFe DOS at the Fermi level is substantially enhanced when the CoFe thickness is reduced from 25 to 15 Å. This enhancement, which we suggest is due to the amorphous character of the CoFe when in thickness and results in a spin-polarized peak in the DOS of primarily Co origin, can be directly correlated to marked improvements in magnetic transport and switching properties. This technique for studying buried-layer DOSs should also be applicable to other multilayer nanostructures.

The structures of the liquid surface of sodium have been characterized with extensive first-principles molecular dynamics simulations based on ensemble density functional theory. Friedel oscillations in the electronic charge density at the free surface were found to persist across the solid-to-liquid melting transition, and remain decoupled from the atomic positions. Strong atomic layering was observed both at the liquid surface and at an artificial liquid–solid interface, notwithstanding the absence of Friedel oscillations or under-coordinated atoms in the latter case. Confinement effects at these soft or hard boundaries drive the atoms into quasi-close-packed layers; even for this prototypical free electron metal Friedel oscillations are not relevant to ordering.

The structure of a straight interface (wall) between regions with differing values of the pitch in planar cholesteric layers with finite strength of the surface anchoring is investigated theoretically. It is found that the shape and strength of the anchoring potential influences essentially the structure of the wall and a motionless wall between thermodynamically stable regions without a singularity in the director distribution in the layer can exist for sufficiently weak anchoring only. More specifically, for the existence of such a wall the dimensionless parameter S_d = K₂₂/Wd (where W is the depth of the anchoring potential, K₂₂ is the elastic twist modulus and d is the layer thickness) should exceed its critical value, which is dependent on the shape of the anchoring potential. General equations describing the director distribution in the wall are presented. Detailed analysis of these equations is carried out for the case of infinitely strong anchoring at one surface and finite anchoring strength at the second layer surface. It is shown that the wall width L is directly dependent upon the shape and strength of the anchoring potential and that its estimate ranges from d to (dL_p)^1/2 (where L_p = K₂₂/W is the penetration length), corresponding to different anchoring strengths and shape potentials. The dependence of the director distribution in the wall upon all three Frank elastic moduli is analytically found for some specific limiting cases of the model anchoring potentials. Motion of the wall is briefly investigated and the corresponding calculations performed under the assumption that the shape of a moving wall is the same as a motionless one. It is noted that experimental investigation of the walls in planar cholesteric layers can be used for the determination of the actual shape of surface anchoring potentials.

We studied the effect of sodium chloride on the phase behaviour of the symmetric triblock copolymer F68 in water using direct visual observation, viscometry, small-angle x-ray scattering and microscopy. We found that salt has a destabilizing effect. The transitions from unimer to micellar solution and from micellar solution to BCC micellar crystal occurred some 20 °C lower in 1 M NaCl than in the salt-free system. Salt also induced a region of liquid–liquid coexistence in the phase diagram at F68 concentrations  wt%, with the lower critical solution temperature (or critical point) occurring at ∼80 °C in 1 M NaCl. At this salt concentration and , liquid–liquid phase separation becomes metastable. A homogeneous micellar solution at this temperature first separated into coexisting dilute and dense micellar liquids, and then the dense liquid would transform into a transparent () and turbid () mesophase. We tentatively identified these as the BCC cubic and lamellar (L_α) phases respectively. A more complex sequence of observations at a Pluronic concentration of 34 wt% confirms this picture. A phase diagram for F68 in 1 M NaCl consistent with all of our observations is proposed, as well as a 'rule of thumb' for interpreting the effect of salt on Pluronic phase behaviour.

Using ab initio molecular dynamics simulations, we have investigated the structure, diffusion and electronic properties of liquid InSb alloy below and above the melting point. A large gap in value or a 'flat shape' between the position of the first and second peak of the calculated pair correlation function, and an asymmetry of the first peak, were displayed in our simulations. There is a small hump between the first and second peak of the partial pair correlation function of Sb atoms, which gradually disappears with increasing temperature. Analysis of the compositional disorder number suggests that heterogeneous bonds still play a predominant role in liquid InSb alloy. With increasing temperature, the opened chain-like Sb–Sb bonds are broken and the Peierls distortion of Sb atoms in liquid InSb alloy is weakened. With the change of temperature, the diffusion coefficients obtained from the average mean-square displacement can be fitted by the Arrhenius equation. The fit yields an activation energy of 0.20 eV and a pre-exponential factor of 2.01 × 10⁻⁸ m² s⁻¹.

For a macroscopic liquid droplet, we applied the fluid mechanics to predict its dynamic spreading where the energies associated with volume are dominant (e.g. Shikhmurzaev 1997 J. Fluid Mech. 334 211). When the droplet size is reduced to the sub-millimetre level, the surface energy becomes the dominant factor. If the droplet size is further reduced to the micrometre or sub-micrometre level, the line tension should be included in our theoretical formulation for predicting droplet spreading. In the present paper, we extend the non-equilibrium thermodynamics (Gao et al 2000 Acta Mater. 48 863–74) to describe the micrometre or nanometre size droplet spreading on a solid substrate with the line tension effect. Discussion of the stable and unstable equilibrium states is made based on the thermodynamic force expression.

A new concentration–temperature phase diagram of superionic Cu₆PS₅Br_1−xI_x mixed crystals with compositions x = 1, 0.75, 0.5, 0.25 is proposed on the basis of calorimetric, optical and single-crystal x-ray diffraction measurements. In the high temperature form, above the first phase transition, all compounds crystallize in cubic symmetry . In iodine rich crystals (for x = 1, 0.75 and 0.5) below 270 K intermediate cubic superstructure is stabilized down to 143–169 K for x = 1 and around 220 K for x = 0.75 and 0.5. Below these temperatures ferroelastic phase transition to monoclinic Cc phase takes place. On the other hand, in bromine rich compounds (x = 0.25) a direct –Cc phase transition is observed without any indications of intermediate cubic superstructure. The influence of compositional disorder on superionic properties is discussed on the basis of conductivity measurements for temperatures between 150 and 290 K. Since conduction properties are coupled with the structure changes the different conductivity behaviour is observed for different I/Br concentration. Ionic conductivity decreases with the increase of bromine concentration. Even in the high temperature phase conductivity is lower and activation energy is higher in bromine rich compounds.

By doping an inorganic compound with two specific types of trivalent rare-earth impurities the controlled creation of electron- and hole-trapping centres is possible. This is demonstrated with experimental data on LiLnSiO₄:Ce³⁺, Sm³⁺, Ln = Y or Lu. After exposure to ionizing radiation electrons are captured by Sm³⁺ and holes are captured by Ce³⁺. The electron trapping depth is given by the energy difference between the Sm²⁺ ground state and the bottom of the conduction band. This energy is estimated from the energy of charge transfer from the valence band to Eu³⁺ employing recently developed models. The trapping energy is also determined from thermoluminescence studies. Both values are in good agreement.

We investigate the dielectric resonance, local field distribution and optical response of randomly ordered nanometre metallic cells in two-dimensional composites using the Green's function formalism. It is found that the microstructure of a single cell determines the bandgap profiles of dielectric resonances, showing that these metallic cells have locally resonant structures. One or several defect bonds are introduced into each cell, expressed as the difference admittance ratio η = (₂−₀)/(₁−₀). With varying η, the resonance bandgaps and effective optical absorption spectra are modified according to the single cell response, or local resonance structure. However, near resonances, the local field distributions from the localized modes to extended modes are shown to be the collective response of the whole material, rather than a single cell response. The results will be beneficial to the study of surface plasmon resonances in nanometre metallic clusters.

We propose an extended Finnis–Sinclair (FS) potential by extending the repulsive term into a sextic polynomial for enhancing the repulsive interaction and adding a quartic term to describe the electronic density function. It turns out that for bcc metals the proposed potential not only overcomes the 'soft' behaviour of the original FS potential, but also performs better than the modified FS one by Ackland et al, and that for fcc metals the proposed potential is able to reproduce the lattice constants, cohesive energies, elastic constant, vacancy formation energies, equations of state, pressure–volume relationships, melting points and melting heats. Moreover, for some fcc–bcc systems, e.g. the Ag–refractory metal systems, the lattice constants, cohesive energies and elastic constants of some alloys are reproduced by the proposed potential and are quite compatible with those directly determined by ab initio calculations.

The binding energies of the ground and some low-lying excited states of a D⁻ centre in a GaAs/GaAlAs quantum well have been calculated as a function of the presence of an applied magnetic field and well width respectively. In comparison with exact two-dimensional results, which show only four bound states, a larger number of D⁻ bound states appear in a quantum well. Moreover, the critical magnetic field values at which the excited states change from unbound to bound are obtained, and the reasons of the binding energy as function of magnetic field and well width are discussed. Our results are in good agreement with those in the literature.

With a reduction in the average grain size in nanostructured films of elemental Nb, we observe a systematic crossover from metallic to weakly insulating behaviour. An analysis of the temperature dependence of the resistivity in the insulating phase clearly indicates the existence of two distinct activation energies corresponding to inter-granular and intra-granular mechanisms of transport. While the high temperature behaviour is dominated by grain boundary scattering of the conduction electrons, the effect of discretization of energy levels due to quantum confinement shows up at low temperatures. We show that the energy barrier at the grain boundary is proportional to the width of the largely disordered inter-granular region, which increases with a decrease in the grain size. For a metal–insulator transition to occur in nano-Nb due to the opening up of an energy gap at the grain boundary, the critical grain size is ≈8 nm and the corresponding grain boundary width is ≈1.1 nm.

The phase relations of the GdCo_13−xSi_x system have been studied by means of scanning microscopy, microprobe analysis and x-ray diffraction. Single-phase samples (structure type LaFe₉Si₄ with space group I4/mcm) are formed in a narrow composition range |δ|≤0.2, where GdCo₉Si₄ forms as a fully ordered ternary compound. The magnetic properties of GdCo₉Si₄ have been investigated by ac susceptibility, magnetization, specific heat and resistivity measurements. These experiments reveal ferrimagnetism below about 47 K, which is analysed in terms of a two-sublattice molecular field model with a local moment Gd subsystem coupling antiparallel to the itinerant ferromagnetic Co 3d sublattice. The 3d–3d exchange of the latter is the driving force for the onset of long range magnetic order. The validity of the two-sublattice model is demonstrated by high field measurements, showing that the ferrimagnetic coupling starts to break up at a lower critical field . The magnetic ground state of GdCo₉Si₄ has been analysed on a microscopic basis via LSDA electronic structure calculations.

We report magneto-transport measurements in parallel magnetic field and μ-Raman spectroscopy on diffusive multiwall carbon nanotubes. The disorder effects on the characteristic transport lengths are probed by combining applied magnetic field and back-gate tuning of the Fermi level. Modulations of the differential conductance versus energy depict the modulation of the strength of the weak localization. Both the electronic mean free path and the phase coherence length are found to be energy dependent. The role of disorder in the density of states and in the characteristic transport lengths is discussed.

Numerical calculations of lattice thermal conductivity are reported for the binary alloys NiPd and NiPt. The present work is a continuation of an earlier paper by us (Alam and Mookerjee 2005 Phys. Rev. B 72 214207), which developed a theoretical framework for the calculation of configuration-averaged lattice thermal conductivity and thermal diffusivity in disordered alloys. The formulation was based on the augmented space theorem (Mookerjee 1973 J. Phys. C: Solid State Phys. 6 L205) combined with a scattering diagram technique. In this paper we shall show the dependence of the lattice thermal conductivity on a series of variables like phonon frequency, temperature and alloy composition. The temperature dependence of κ(T) and its relation to the measured thermal conductivity is discussed. The concentration dependence of κ appears to justify the notion of a minimum thermal conductivity as discussed by Kittel, Slack and others (Kittel 1948 Phys. Rev. 75 972, Brich and Clark 1940 Am. J. Sci. 238 613; Slack 1979 Solid State Physics vol 34, ed H Ehrenreich, F Seitz and D Turnbull (New York: Academic) p 1). We also study the frequency and composition dependence of the thermal diffusivity averaged over modes. A numerical estimate of this quantity gives an idea about the location of the mobility edge and the fraction of states in the frequency spectrum which is delocalized.

The electrical and photoelectrical properties of In₆S₇ crystals have been investigated in the temperature regions of 170–300 K and 150–300 K, respectively. The dark electrical analysis revealed the intrinsic type of conduction. The energy band gap obtained from the temperature-dependent dark current is found to be 0.75 eV. It is observed that the photocurrent increases in the temperature range of 150 K up to T_m = 230 K and decreases at T>T_m. Two photoconductivity activation energies of 0.21 and 0.10 eV were determined for the temperature ranges below and above T_m, respectively. The photocurrent (I_ph)–illumination intensity (F) dependence follows the law . The value of γ decreases when the temperature is raised to T_m, then it starts increasing. The change in the value γ with temperature is attributed to the exchange in role between the recombination and trapping centres in the crystal. The crystals are found to exhibit photovoltaic properties. The photovoltage is recorded as a function of illumination intensity at room temperature. The maximum open-circuit voltage and short-circuit photocurrent density, which are related to an illumination intensity equivalent to one sun, are 0.12 V and 0.38 mA cm⁻², respectively.

Considerable efforts have been devoted recently to synthesizing diluted magnetic semiconductors having ferromagnetic properties at room temperature because of their technological impacts for spintronic devices. In 2001 successful growth of GaMnN films showing room temperature ferromagnetism and p-type conductivity was reported. The estimated Curie temperature was 940 K at 5.7% of Mn, which is the highest among diluted magnetic semiconductors ever reported. However, the electronic mechanism behind the ferromagnetic behaviour has still been controversial. Here we show experimental evidence using ferromagnetic samples that Mn atoms are substitutionally dissolved into the GaN lattice and they exhibit mixed valences of +2 (majority) and +3. The p-type carrier density decreases significantly at very low temperatures. At the same time, magnetization dramatically decreases. The results imply that the ferromagnetic coupling between Mn atoms is mediated by holes in the mid-gap Mn band.

A first-principles procedure for the calculation of equilibrium properties of crystals under hydrostatic pressure is applied to Ca, Sr and Ba. The procedure is based on minimizing the Gibbs free energy G (at zero temperature) with respect to the structure at a given pressure p, and hence does not require the equation of state to fix the pressure. The calculated lattice constants of Ca, Sr and Ba are shown to be generally closer to measured values than previous calculations using other procedures. In particular for Ba, where careful and extensive pressure data are available, the calculated lattice parameters fit measurements to about 1% in three different phases, both cubic and hexagonal. Rigid-lattice transition pressures between phases which come directly from the crossing of G(p) curves are not close to measured transition pressures. One reason is the need to include zero-point energy (ZPE) of vibration in G. The ZPE of cubic phases is calculated with a generalized Debye approximation and applied to Ca and Sr, where it produces significant shifts in transition pressures. An extensive tabulation is given of structural parameters and elastic constants from the literature, including both theoretical and experimental results.

A phenomenological approach is presented for the giant magnetoresistance (GMR) of magnetic multilayers oriented in the CPP mode (current perpendicular to the plane of the layers). New results are found for the dependence of the GMR on the number of repeats of the multilayer.

We have performed an ab initio study of the energetics, structural, electronic, and optical properties of the CH₄-doped ultrathin 4 Å and large diameter carbon nanotubes (CNTs). No adsorption of the CH₄ molecule has been seen either on the groove or on the interstitial sites of the achiral 4 Å nanoropes. In the case of weakly bonded systems such as the adsorption of the CH₄ molecules on the nanotubes, a prominent role of the dispersion forces in the binding is observed. The negative contribution of the zero point vibrational energy to the binding energy is seen to be appreciable. The effects of the tube diameter and the different chiralities of the carbon nanotubes on the adsorbate-induced physical properties have been investigated. In the large diameter tubes, the binding of the CH₄ molecule in the endohedral adsorption is much stronger than that in the exohedral adsorption. The binding of the CH₄ molecules depends upon the chirality of the nanotube and we find no adsorption on the chiral (4, 2) tube. We find that the local density approximation (LDA) over-binds the CH₄ molecule and the generalized gradient approximation (GGA) under-binds it, and for a reliable theoretical estimate one should take some weighted average of the binding energies (BEs) determined in the LDA and the GGA. The currently calculated BE and the adsorbate concentration are in reasonable agreement with the measured data available for the (10, 10) nanotubes. The electronic structure of the pristine tube is quite altered by the adsorption of the CH₄ molecule on the surface of the tube because of the breaking of the symmetry of the host lattice except the chiral (4, 2) tube, which has practically no symmetry. The adsorption incurs splitting in the states in the whole energy range, especially in the large curvature 4 Å tubes. The bandgap of the semiconducting achiral zigzag nanotube is reduced, whereas that of a chiral semiconducting tube is enhanced, by the adsorption of the CH₄ molecules. The adsorption of CH₄ molecules does not alter significantly the peak structure in the optical absorption of the pristine tube, except for some changes in the energy locations and the relative intensities in the achiral tubes. Most of the calculated peaks in the optical absorption of the pristine large diameter (10, 0) and (10, 10) nanotubes have been observed in the experimental measurements.

We report on measurements of the temperature dependent specific heat, resistivity, Seebeck and Hall coefficients of U₂Rh₂In. This compound does not order magnetically down to 1.5 K and is characterized by an enhanced electronic-heat coefficient γ≈130 mJ/mol U K² at 2 K. The magnetic resistivity shows a T^3/2 law below 4 K and a broad maximum at T_max^ρ = 55 K. The latter refers to the onset of coherence between U ions. The application of pressures up to 19 kbar diminishes the contribution of the spin fluctuation to the resistivity, but does not significantly change the position of the resistivity maximum. Thermopower is characterized by a deep minimum of −30 µV K⁻¹ at 10 K, followed by a sign change at 45 K. The Hall coefficient shows a maximum at ∼T_max^ρ and is dominated by skew scattering. At 2 K, the ordinary Hall coefficient R₀≈9.65 × 10⁻¹⁰ m³ C⁻¹ corresponds in a one-band model to a charge carrier density of 1.33 holes/formula unit. The effective mass of the carriers at low temperatures is estimated to be 55m_e.

Nanocrystalline Ni–NiAl₂O₄–YSZ cermet with a possible application as anode in solid oxide fuel cells (SOFCs) has been developed. The powders were prepared by using an alternative solid-state method that includes the use of nickel acetylacetonate as an inorganic precursor to obtain a highly porous material after sintering at 1400 °C and oxide reduction () at 800 °C for 8 h in a tubular reactor furnace using 10% H₂/N₂. Eight samples with 45% Ni and 55% Al₂O₃–YSZ in concentrations of Al₂O₃ oxides from 10 to 80 wt% of were mixed to obtain the cermets. The obtained material was compressed using unidirectional axial pressing and calcinations from room temperature to 800 °C. Good results were registered using a heating rate of 1 °C min⁻¹ and a special ramp to avoid anode cracking. Thermal expansion, electrical conductivity, and structural characterization by thermo-mechanical analyser (TMA) techniques/methods, the four-point probe method for conductivity, scanning electron microscopy (SEM), x-ray energy dispersive spectroscopy (EDS), x-ray diffraction (XRD), and the Rietveld method were carried out. Cermets in the range 5.5 to 11% Al₂O₃ present a crystal size around 200 nm. An inversion degree (I) in the NiAl₂O₄ spinel structure of the cermets Ni–NiAl₂O₄–YSZ was found after the sintering and reduction processes. Good electrical conductivity and thermal expansion coefficient were obtained for the cermet with 12 wt% of spinel structure formation.

Aluminate glasses are difficult to prepare as they do not contain traditional network formers, but they are promising materials for optical applications. The atomic structure of calcium aluminate glasses has been studied using several experimental techniques. The current study uses molecular dynamics to obtain a model of a (CaO)_0.625(Al₂O₃)_0.375 glass close to the eutectic. The glass consists of a tetrahedral alumina network with average network polymerization of n = 3.3. Ca acts as a network modifier with average coordination of 6.2. Ca is typically coordinated to three bridging oxygens (O_b) and three non-bridging oxygens (O_nb), with Ca–O_nb bonds noticeably shorter than the Ca–O_b bonds. A new method of analysing modifier cation coordination is presented, which specifically shows the distribution of Ca coordination N_CaO in terms of combinations of N_CaOb and N_CaOnb. O_b is most often coordinated to two Al plus two Ca, and O_nb is most often coordinated to one Al plus three Ca. The typical coordinations of Ca, O_b, and O_nb all have a noticeable similarity to those for the 5CaO·3Al₂O₃ crystal. The Ca–Ca distribution shows a clear similarity to that for (CaO)_0.5(SiO₂)_0.5 glass, and this is attributed to the equal atomic number densities of Ca in these glasses.

Perovskite Ba_0.5Sr_0.5TiO₃ (BST) thin films, with thickness of about 350 nm, have been epitaxially grown on (001) LaAlO₃ substrates by pulsed-laser deposition. The good crystallography and epitaxy characteristics were confirmed using x-ray diffraction and transmission electron microscopy (TEM). The dielectric properties of the BST thin films were measured with a planar capacitor configuration in the temperature range of 77–300 K. The capacitance–temperature characteristics, measured with no and several different DC biases, reveal that the BST films are in the paraelectric state at room temperature, with a very good dielectric tunability. However, a butterfly-shaped curve, typical for a ferroelectric material, was obtained from room temperature capacitance–voltage (C–V) measurements, suggesting a faint ferroelectric-like effect for the BST thin films. Careful dielectric property characterizations showed significant temperature and frequency dependence, probably indicating a behaviour of Maxwell–Wagner-type dielectric relaxation. As a result, the weak C–V hysteresis effect in our paraelectric BST thin films is believed to be ascribable both to oxygen vacancies and to the presence of other space charges, trapped at the grain boundaries and/or at the substrate/dielectric film interface, which give rise to local polar regions in the thin-film samples. Furthermore, this explanation is supported by cross-sectional TEM and off-axis electron holographic observations.

A quasi-two-dimensional antiferromagnet with a square-lattice bilayer structure is studied with a bond operator formalism. The excitation spectrum and its temperature dependence, the quantum phase transition from the spin-singlet dimerized state to the Néel state, are studied. In particular the field-induced long range order is interpreted using Bose–Einstein condensation of magnons. The phase diagram as well as the temperature and magnetic field dependences of the specific heat and the staggered magnetization are studied. The critical exponents around the critical magnetic fields H_{c 1} and H_{c 2} are calculated to be about 1.5. These results agree well with recent experimental results for BaCuSi₂O₆.

We have measured the magnetic field (H≤90 kOe) and pressure (P≤10 kbar) dependence of the magnetic ordering temperature, T_mag, in single crystal samples of Na_xCoO₂ for a range of Na concentrations (0.60≤x≤0.72). We show that in zero field T_mag remains constant with decreasing x before magnetic order disappears at x = 0.65. Heat capacity and magnetization data show that for x = 0.70 T_mag is unchanged in an applied field. In contrast, magnetization data collected under hydrostatic pressure show that T_mag increases from 22.0 K at 1 bar to 25.4 K at 10 kbar. This rise is at odds with the behaviour expected for model spin density wave systems.

Scattering of neutrons in the 24–150 keV incident energy range from H₂O relative to that of D₂O and H₂O–D₂O mixtures was reported recently by Moreh et al. This work is related to neutron Compton scattering experiments regarding the 'anomalous' scattering from protons, observed earlier at ISIS by Chatzidimitriou-Dreismann et al in the 5–100 eV range. Here we provide the complete data reduction scheme of time-of-flight integrated intensities measured at keV energy transfers, within the impulse approximation of standard theory and for single scattering events. Current investigations of multiple scattering events and the associated preliminary results are mentioned. Direct application of the theoretical results to the new keV scattering data reveals an anomalous ratio of scattering intensity of H₂O relative to that of D₂O of about 20%, thus being in good agreement with the earlier results of the original experiment at ISIS.

The results of X-band electron paramagnetic resonance (EPR) measurements of Co²⁺ ions in YAlO₃ (YAP) crystals in the temperature range 1.8–40 K are presented. The temperature and angular dependences of EPR spectra have been analysed using a triclinic spin Hamiltonian (SH) consisting of the electronic Zeeman and hyperfine terms. Two distinct positions α and β are identified for Co²⁺ complexes and ascribed to the substitutional Co²⁺ ions at the Al³⁺ and Y³⁺ sites, respectively. The values of the SH parameters are obtained by least squares fitting the α- and β-type Co²⁺ spectra yielding the principal (orthorhombic-like) values of the tensors g and A as well as the orientation of their principal axes. The additionally observed EPR spectra of the unintentional impurities Nd³⁺ and Er³⁺ in YAP crystals are also analysed.

Low-frequency Raman spectra of densified silica glasses, annealed at different temperatures and possessing various densities, are studied. The dependence of the fast relaxation intensity versus the density can be described by linear laws below and above the density 2.26 g cm⁻³, but the slope of the linear law below the critical density is ten times higher than the slope of the linear law above 2.26 g cm⁻³. The sharp peculiarity is interpreted as a consequence of an inhomogeneous glassy structure with local ordering similar to tridimite or β-cristobalite. The dependence of the silica sample density on annealing temperature is discussed in view of the temperature dependence of the liquid silica density. A hypothesis is proposed that the true dependence of the liquid silica density is a linear function of temperature.