Photoemission spectra of single crystal ${\mathrm{CeSi}}_{\mathit{x}}$ (x=1.86; nonmagnetic, ${\mathit{T}}_{\mathit{K}}$∼20 K, and x=1.7; ferromagnetic, ${\mathit{T}}_{\mathit{c}}$=13.7 K) were measured at very high-energy resolution and low temperature. The Fermi-edge band in the photoemission spectrum of ${\mathrm{CeSi}}_{1.86}$ is narrower and closer to ${\mathit{E}}_{\mathit{F}}$ compared with the previous reports on polycrystals, showing closer resemblance to the calculation based on the Kondo model. The intensity of the Fermi-level peak is drastically reduced in going from ${\mathrm{CeSi}}_{1.86}$ to ${\mathrm{CeSi}}_{1.7}$, which is in good qualitative agreement with the Kondo model.

Hydrogen insertion in the ${\mathrm{Ce}}_2$${\mathrm{Fe}}_{17}$${\mathrm{H}}_{\mathit{x}}$ system leads to spectacular changes of the physical properties. A relative increase of the Curie temperature of almost 100% is observed as well as the highest volume increase. In addition to previous neutron diffraction measurements which pointed out an increase of the local iron moments, we present in this paper absorption and magnetic circular x-ray dichroism (MCXD) experiments performed at the ${\mathit{L}}_{2,3}$ edges of Ce and the K edge of iron. These measurements allow us to study the evolution of the Ce electronic structure upon hydrogen insertion. Values for the Ce valence have been derived from both absorption and MCXD signals. The MCXD data are analyzed for their shape, intensity, and sign, and compared to the macroscopic magnetic measurements. It is shown that in these compounds the Ce atoms bear a 5d magnetic moment, which is coupled antiferromagnetically to that on iron. There is no orbital contribution to this 5d moment which is thus solely of spin origin. The values of the 5d moment are deduced as a function of each composition in the ${\mathrm{Ce}}_2$${\mathrm{Fe}}_{17}$${\mathrm{H}}_{\mathit{x}}$ system at low and room temperature. At the K edge of iron, the shape and intensity of the dichroic signal have been related to the filling of the majority 3d band.

The Bloch equations with radiation damping terms are solved. Analytical results are obtained for the case where the longitudinal relaxation effect can be neglected. General line-shape expressions, which can be reduced to those in conventional NMR, are presented. The line-shape distortions of the NMR signals due to radiation damping effects are analyzed. Unusual intensity profiles as a function of the pulse-flip angle ${\mathrm{\theta }}_0$, which are strongly dependent on the ratio of ${\mathit{T}}_2^{\mathrm{*}}$/${\mathit{T}}_{\mathit{r}}$ (where ${\mathit{T}}_2^{\mathrm{*}}$ and ${\mathit{T}}_{\mathit{r}}$ are the effective transverse relaxation time and the radiation damping time, respectively), are predicted. Radiation damping effects on the phase and the linewidth of the signals are also discussed. Some experimental results are given.

In this work, we have studied the invariants of motion of two SU(N) supersymmetric t-J models of 1/${\mathit{r}}^2$ hopping and exchange in one dimension. The first model is defined on a lattice of equal spaced sites, and the second on a nonequal spacing lattice. Using the ‘‘exchange operator formalism,’’ we are able to construct all the invariants for the models, by mapping the systems onto mixtures of fermions and bosons. This identification shows that the supersymmetric t-J model on the chain with equal-spaced sites also belongs to Shastry-Sutherland’s ‘‘super-lax-pair’’ family.

To understand the relationship between the structure and magnetism of fcc Pd-Mn spin-glass disordered alloys, we have measured the lattice parameters, electron diffraction (ED) patterns, x-ray diffuse scattering intensity, and magnetic susceptibilities. The lattice parameters of the ${\mathrm{Pd}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Mn}}_{\mathit{x}}$ alloys show a nonlinear increase with composition for 0≤x≤0.25. Peculiar diffuse scattering has been observed at the 100, 110, and the equivalent positions on ED patterns in the composition range from x=0.15 to 0.25. Warren-Cowley short-range order (SRO) parameters from a single crystal of the x=0.25 alloy were obtained by analyzing the x-ray diffuse intensity. Extra diffuse maxima were observed at 1/2,1/2,1/2 and equivalent positions, in addition to the diffuse scattering described above, and their intensity is 20% of that at the 100 position. A correlation length has been deduced to be about 10 Å from estimating the reconstructed diffuse intensity distribution using the SRO parameters determined experimentally. A magnetic phase diagram has been constructed from measuring the freezing temperature (${\mathit{T}}_{\mathit{g}}$). ${\mathit{T}}_{\mathit{g}}$ shows a peculiar dependence for a Mn content below x=0.05, above which it increases monotonically up to x=0.17, and then rises dramatically up to x=0.25. The paramagnetic Curie temperature (FTHETA) is negative for all the compositions, which indicates that the antiferromagnetic interactions are influential. Ordered-structure experiments indicate that the values of ${\mathit{T}}_{\mathit{c}}$ are very low for alloy compositions of x<0.244.

The volume dependence of magnetic properties in amorphous Fe has been calculated on the basis of the finite-temperature theory of amorphous metallic magnetism which self-consistently determines the distribution of local magnetic moments. The calculated magnetic phase diagram on the T-V plane and magnetization-vs-volume curve show the existence of the spin-glass state in a wide range of volume (10.50 A${\mathrm{̊}}^3$≲V≲12.5 A${\mathrm{̊}}^3$) after disappearance of ferromagnetism. The results verify our previous conclusion of the spin-glass state in amorphous Fe, which is expected from the experimental data of Fe-rich amorphous alloys containing early transition metals and rare-earth metals, but disagree with those obtained from the supercell approaches in the ground-state electronic structure calculations. It is shown that the nonlinear magnetic couplings between the nearest-neighbor Fe local moments and the local environment effects on the amplitude of Fe local moments via structural disorder lead to the spin-glass state in the weak magnetic region, while the volume expansion develops the ferromagnetic couplings and therefore the ferromagnetism.

The short-ranged correlations associated with magnetic ordering in the rare-earth antiferromagnet holmium have been characterized in high-resolution x-ray- and neutron-scattering studies. For temperatures within about 2 K above ${\mathit{T}}_{\mathit{N}}$, the short-ranged magnetic fluctuations exhibit two length scales, instead of the single one expected in an ideal system. Well above ${\mathit{T}}_{\mathit{N}}$, the shorter of the two length scales exhibits power-law behavior consistent with normal critical fluctuations; the line shape in momentum space is well described by a Lorentzian, and the measured critical exponents are ν=0.54±0.04 and γ=1.24±0.15. The longer of the two length scales is well described by a squared-Lorentzian line shape, and exhibits a power-law temperature dependence. Both the shorter and the longer length-scale fluctuations are approximately spatially isotropic. We propose that the longer length-scale fluctuations are related to random strain fields which are localized at or near the sample surfae. These results are reminiscent of behavior observed at the cubic-to-tetragonal structural phase transitions of some perovskite materials.

A two-dimensional, numerical, micromagnetic model was applied to understand the images of 180° walls in iron thin films obtained with a magnetic force microscope (MFM). Assuming the tip stray field would not affect the wall magnetization, the model predicts a contrast well below the sensitivity of the MFM. However, this assumption is unrealistic. The inclusion of the tip stray field in the simulation shows that the domain wall is polarized by the tip stray field. The resulting contrasts from the polarized walls differ drastically from the contrasts expected from undisturbed walls but agree well with experimental observations. Such polarization effects make the authors doubtful about the applicability of the MFM to studies of other micromagnetic phenomena in soft magnetic materials.

Motivated by difficulties in understanding the magnetism of the insulating parent of high-${\mathit{T}}_{\mathit{c}}$ supeconductors, we have studied the less covalent, and thus simpler, antiferromagnetic (AF) insulators NiO and ${\mathrm{KNiF}}_3$. We also consider the apparently covalent material ${\mathrm{La}}_2$${\mathrm{NiO}}_4$, which is closely related to the high-${\mathit{T}}_{\mathit{c}}$ superconductor parent ${\mathrm{La}}_2$${\mathrm{CuO}}_4$. Despite many studies of ${\mathrm{KNiF}}_3$ and NiO via cluster calculations, we found that a satisfactory ab initio cluster theory of the neutron form factor is lacking. We have carried out such a calculation in the unrestricted Hartree-Fock (UHF) approximation, taking the basic cluster as (${\mathrm{NiF}}_6$${)}^{4\mathrm{-}}$ and (${\mathrm{NiO}}_6$${)}^{10\mathrm{-}}$ for ${\mathrm{KNiF}}_3$ and NiO, respectively, treating the remaining lattice in the point-charge model. We show that correlation effects and Pauli repulsion corrections to the point charges are negligible in these cases. After correcting for the zero-point spin fluctuations, the UHF form factor agrees well with experiment in ${\mathrm{KNiF}}_3$, where the absolute value of the form factor is known for small scattering vectors q. The UHF calculations agree satisfactorily with the relative form factor data for NiO, which cover a large range of ‖q‖ (the absolute experimental values are not available); the agreement in shape includes the variations with q due to asphericity of the spin density. We also found that the UHF results on ${\mathrm{La}}_2$${\mathrm{NiO}}_4$, obtained by using similar methods, disagree sharply with experiment.

The long-range planar-type antiferromagnetic ordering in the heavy-fermion compound ${\mathrm{CePd}}_2$${\mathrm{Al}}_3$ is shown to critically depend on the exact occupation and distribution of aluminum atoms over their available lattice sites. A poly-crystalline sample (${\mathit{T}}_{\mathit{N}}$=2.7 K) is systematically compared with a nonordering single crystal, by means of susceptibility, specific heat, and site-specific Al-NQR/NMR experiments. Implications for other Al-containing heavy-fermion systems are discussed.

A Monte Carlo study of the two-dimensional randomly site-diluted Ising system is reported. The data reveal that the specific heat does not diverge at the critical point, at least when the concentration of dilution is sufficiently large. The critical behavior of the magnetic susceptibility and the correlation length are consistent with pure power-law behavior with concentration-dependent critical indices.

The S=1 antiferromagnetic Heisenberg chain both with and without single-ion anisotropy is studied. Using the recently proposed density-matrix renormalization-group technique of White, we calculate the energy gaps as well as several different correlation functions. The two gaps, ${\mathrm{\Delta }}_{\mathrm{\parallel }}$ and ${\mathrm{\Delta }}_{\mathrm{\perp }}$, along with associated correlation lengths and velocities are determined. The numerical results are shown to be in good agreement with theoretical predictions derived from the nonlinear σ model and a free-boson model. We also study the S=1/2 excitations that occur at the ends of open chains; in particular we study the behavior associated with open boundary conditions, using a model of S=1/2 spins coupled to the free bosons.

We have used the technique of polarized-neutron reflectometry to study the magnetic-induction profile just beneath the surface of a 1-μm-thick film of lead with the magnetic field applied parallel to the surface. The sample was maintained at a temperature of 1.5 K throughout the experiment, and the applied field H was varied from below ${\mathit{H}}_{\mathit{c}}$, the bulk critical field, up to the critical field for surface superconductivity, ${\mathit{H}}_{\mathit{c}3}$. From the measurements with the lead film in the bulk superconducting phase (H<${\mathit{H}}_{\mathit{c}}$), we found that the spin-dependent reflectivity profiles are consistent with an exponential decay of magnetic induction with a penetration depth λ=39±1 nm. At higher applied fields (${\mathit{H}}_{\mathit{c}}$<H<${\mathit{H}}_{\mathit{c}3}$) we obtained information on the diamagnetism in the surface superconducting layer. We discuss the sensitivity of the measurements to the form of the magnetic-induction profile, in particular to nonlocal effects, and show that a good description of the surface diamagnetism, over the whole range of applied fields, can be achieved with the local Ginzburg-Landau theory if the Ginzburg-Landau parameter κ is allowed to vary with the applied field.

The proclivity of paramagnetic solutes to degrade the transition temperature of a traditional singlet-pairing superconductor is studied anew. ${\mathit{T}}_{\mathit{c}}$ degradation is proportional to the (conduction electron) spin-disorder scattering rate, 1/${\mathrm{\tau }}_{\mathit{s}}$, caused by the solute spins. Accordingly the ${\mathit{T}}_{\mathit{c}}$ loss increases with solute concentration. The initial slope (versus 1/${\mathrm{\tau }}_{\mathit{s}}$) is found to depend on the superconductor and therefore is not the universal constant proposed by Abrikosov and Gor’kov. Instead the decrease is inversely proportional to λ, the electron-phonon coupling constant. Consequently a weak superconductor is doubly jeopardized by paramagnetic impurities: Its superconductivity is easily suppressed not only because ${\mathit{T}}_{\mathit{c}}$ is small to begin with, but also because the initial slope is steeper. Another unexpected consequence of the theory involves potential scattering which, acting alone, does not significantly influence ${\mathit{T}}_{\mathit{c}}$ (as surmised by Anderson). Nevertheless, the ${\mathit{T}}_{\mathit{c}}$ reduction caused by exchange scattering will be partially suppressed when the overall mean free path becomes smaller than the coherence length. This compensation has been demonstrated experimentally by comparing the influence of magnetic impurities in a pure host superconductor with that in a similar host having (also) nonmagnetic solutes. Such observed recovery of ${\mathit{T}}_{\mathit{c}}$, expected from this study, contradicts prior theories for magnetic solutes.

The puzzling problem of flux-flow noise is investigated both theoretically and experimentally. Combined measurements of the voltage noise and magnetic flux noise in various search coils have been performed, using the classical geometry of soft long strips in a perpendicular field. Our samples are immersed in superfluid helium (T∼1.8 K), allowing one to widely explore the flux-flow regime. Auto and cross spectra, in the 1–${10}^4$ Hz range, provide decisive information about the possible sources of noise. In particular, experimental results cannot be reconciled with a two-dimensional (2D) picture of the vortex motion, especially with the classical shot-noise model. A recently presented 3D mechanism of noise [Phys. Rev. Lett. 70, 1521 (1993)] is discussed at greater length in this paper. According to a continuum theory of transport in type-II superconductors by two of us (P.M. and Y.S.), the transport current J is separated into a nondissipative surface current ${\mathbf{J}}_1$ and a bulk dissipative current ${\mathbf{J}}_2$. The former is associated with the instantaneous distorted configuration of the vortex array, whereas the latter is connected with its motion. A model is developed, that describes surface currents as a 2D homogeneous turbulent flow. This is characterized locally by its normalized spectrum Σ(f), the intensity of the turbulence—denoted as ${\mathit{u}}^{\mathrm{*}}$(∼${10}^2$ A/m) in the text, and directly related to the critical current density—and a correlation length c∼1–3 μm. Auto or cross spectra of both voltage and magnetic field noises can be entirely predicted as a function of Σ(f), ${\mathit{u}}^{\mathrm{*}}$, and c, in full agreement with experiment.

Results of full-potential linear-muffin-tin-orbital calculations for the new intercalated niobium oxide phases ${\mathrm{Li}}_{\mathit{x}}$${\mathrm{NbO}}_2$ and ${\mathrm{Na}}_{\mathit{x}}$${\mathrm{NbO}}_2$ are presented. It is shown that alkali atoms are ionized, and their valence electrons completely fill the hybridized Nb d–O 2p conduction band in ${\mathit{M}}_{1.0}$${\mathrm{NbO}}_2$ crystals making them semiconducting. In the nonstoichiometric ${\mathit{M}}_{\mathit{x}}$${\mathrm{NbO}}_2$ phases, the Fermi level is located well inside the conduction band of predominant Nb 4d nature with significant contributions from O 2p states. Estimates of electron-phonon interaction parameters and ${\mathit{T}}_{\mathit{c}}$ in the scope of the rigid-ion approximation result in ${\mathit{T}}_{\mathit{c}}$∼12 K for ${\mathrm{Li}}_{0.5}$${\mathrm{NbO}}_2$ and ${\mathit{T}}_{\mathit{c}}$∼9 K for ${\mathrm{Na}}_{\mathit{x}}$${\mathrm{NbO}}_2$ in reasonable agreement with the experimental data—and show that superconductivity observed in these materials is of the usual electron-phonon type, similar to that in cubic ${\mathrm{Li}}_{0.5}$${\mathrm{TiO}}_2$. It is shown that the density of states at the Fermi level and ${\mathit{T}}_{\mathit{c}}$ for ${\mathrm{Na}}_{\mathit{x}}$${\mathrm{NbO}}_2$ should be more critically dependent on stoichiometry than for ${\mathrm{Li}}_{\mathit{x}}$${\mathrm{NbO}}_2$. Since the electronic-structure peculiarities obtained in the calculations are very different than those typical for superconducting cuprates, it does not appear likely that any high-${\mathit{T}}_{\mathit{c}}$ superconductors should be expected in this class of compounds.

This method of correlated basis functions is applied at the variational level to give an optimized description, at zero temperature, of the structure and elementary excitations of liquid ${}_{}{}^4{}_{}{}^{}\mathrm{He}$ in the geometry of a half-space. A trial ground-state wave function of Hartree-Jastrow form is assumed, and the Feynman ansatz is adapted to construct trial elementary excitations based on this variational ground state. Functional variation of the energy expectation value with respect to ground and excited trial states leads, in conjunction with the Bogoliubov-Born-Green-Kirkwood-Yvon relations and the hypernetted-chain (HNC) equations, to coupled Euler-Lagrange equations consisting of (i) a modified Hartree equation (ii), a paired-phonon equation, and (iii) a renormalized Bogoliubov eigenvalue equation. These relations and equations provide for simultaneous optimal determination of (i) the density profile, the chemical potential, and the Hartree inhomogeneity factor, (ii) the anisotropic two-body pseudopotential and two-body spatial distribution function, and (iii) the wave functions and energies of the Feynman excitations as functions of the momentum parallel to the surface plane. In the numerical calculation reported, the bulk liquid density is taken equal to the experimental value at saturation. Since the corresponding Jastrow variational treatment of the bulk liquid does not produce a self-bound system at this density, an external potential is introduced to stabilize the surface, its strength being adjusted so that the calculated chemical potential matches the experimental saturation value.

The calculation yields dispersion relations for two distinct branches of bound surface states, extending from the continuum of liquid states at small wave numbers to the continuum of liquid states close to the wave number characteristic of a bulk roton. The two branches are distinguished by the number of nodes (zero or one) of the corresponding wave functions in the surface region. At small wave numbers, the wave functions of the lowest-lying surface states penetrate exponentially into the bulk liquid to a characteristic depth proportional to wavelength. These modes are associated with surface phonons and capillary waves, being driven by the external potential (renormalized by correlation effects due to the strong internal forces) and by the surface tension. The spectrum of surface excitations of the first branch follows the hydrodynamic dispersion relation in the range of wave numbers 0≤q≤0.5 A${\mathrm{̊}}^{\mathrm{-}1}$. Employing a specialized version of the renormalized Bogoliubov equation, analytic expressions are derived that permit evaluation of the speed of surface sound and the surface-tension coefficient in terms of quantities generated by the microscopic calculation. In the opposite regime of large wave numbers corresponding to the atomic scale, q≥1 A${\mathrm{̊}}^{\mathrm{-}1}$, the wave functions of the first branch are centered at a local density approaching that of the bulk liquid. The dispersion curves of both branches appear to terminate by merging with the bulk excitation curve near the roton minimum, in conformity with the interpretation of the bound surface states in this wave-number range as trapped rotons.

Metallic κ-(BEDT-TTF${)}_2$Cu[N(${\mathrm{CN}}_2$]Cl turns semiconducting and insulating at low temperatures and ambient pressure and enters a ferrimagneticlike state below 20 K. Under hydrostatic pressure (p≃300 bar), previous investigations have shown that the metallic state can be stabilized and that reentrant superconducting behavior can be observed (by means of resistivity measurements) in a rather wide pressure range. By means of dc susceptibility measurements, we show that reentrant volume superconductivity actually exists only in a narrow critical pressure range. In contrast to previously found reentrant superconductors, this material does not contain rare-earth magnetic moments.

We study the Josephson current through a ballistic, normal metal, mesoscopic, one-dimensional quantum channel in contact with two superconducting electrodes. The two superconductors have different energy gaps ${\mathrm{\Delta }}_1$ and ${\mathrm{\Delta }}_2$. Increasing either energy gap increases the critical current ${\mathit{I}}_{\mathit{c}}$. This enhancement of ${\mathit{I}}_{\mathit{c}}$ results from altering the mesoscopic processes which control the Josephson-current flow, rather than through the proximity effect. The effect is strongest in junctions shorter than the size of a Cooper pair. When the asymmetry is large, such that ${\mathrm{\Delta }}_2$≫${\mathrm{\Delta }}_1$, we find the critical current can be several times larger than its bulk value in the weaker superconductor, namely, ${\mathit{I}}_{\mathit{c}}$>e${\mathrm{\Delta }}_1$/ħ.

The order parameter and exchange field of superfluid ${}_{}{}^3{}_{}{}^{}\mathit{B}$ near magnetically active surfaces with perfect reflection are calculated using the quasiclassical theory and a magnetically scattering surface model. Our results show that an exchange field is usually induced in the vicinity of such surfaces due to the spin rotation of the impending quasiparticles, but for certain magnetic configuration of the surface, the order parameters are not depressed at all. This is in strong contrast to the situation near nonmagnetic surfaces where the perpendicular order parameter always vanishes due to the triplet pairing nature of the Cooper pair. We discuss the possible application of this property to identify the Cooper pair symmetry of other unconventional superconductors.