We have measured the differential conductance of two different one-dimensional (1D) constrictions with several occupied subbands at low temperature under conditions of high dc source-drain bias relative to the 1D subband spacing. The results are compared with a simulation based on a saddle-point potential with and without a thin equipotential strip at the center. We fit the experimental results at high conductance by increasing the width of the equipotential strip with an increase in number of conducting subbands. For a split gate where the depth of the electron gas is 70 nm and where there are at most three occupied 1D subbands, a simple parabolic saddle-point model is sufficient. For a device where the depth of the electron gas is 27 nm and there are eight occupied 1D subbands, the modified potential model is needed to give good agreement with experimental data.

Photoluminescence studies of GaAs/${\mathrm{Al}}_{\mathit{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathit{x}}$As and ${\mathrm{In}}_{\mathit{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathit{x}}$As/GaAs quantum-well wires have been performed. The emission from the GaAs wires is dominated by a spatially indirect transition, and the ${\mathrm{In}}_{\mathit{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathit{x}}$As wires by a direct transition. Detailed time-resolved-photoluminescence studies show a redshift in photoluminescence energy with increasing time delay after excitation for the indirect transition in the GaAs wires, accompanied by a nonexponential decay time which increases with increasing delay. The ${\mathrm{In}}_{\mathit{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathit{x}}$As wires show no such behavior, with a decay time constant that is independent of time delay after excitation.

The existence of an image charge causes a modification of the tunnel barrier shape, which depends on the effective junction capacitance C. The tunneling rate can be calculated using the expression of the orthodox theory of single-electron tunneling, with an additional prefactor of the order of exp(τ${\mathit{e}}^2$/ħC) where τ is the tunneling traversal time.

The atomic and magnetic structure of fcc Fe grown on Cu(100) are determined using spin-density-functional theory and the full-potential linear muffin-tin orbital method. Different types of interlayer and intralayer magnetism are investigated. For films of four up to eleven adsorbed Fe layers our ab initio total-energy calculations always give a ferromagnetic intralayer spin ordering of the Fe atoms. The surface and first subsurface layers couple ferromagnetically, whereas the deeper ones show interlayer antiferromagnetic coupling. This magnetic ordering is accompanied by a 3.9% expansion of the first layer spacing and a 1% contraction of the second, which agree well with low-energy electron diffraction analyses. We link the results to ground-state calculations for fcc iron.

A tensor LEED analysis of the Cu(110)-(2×3)-N surface structure supports a reconstruction model in which the topmost layer can be described as a pseudo-(100)-c(2×2)-N overlayer, with metal corrugation of about 0.52 Å in the reconstructed layer. Each N atom is nearly coplanar with the local plane formed by the four neighboring Cu atoms. Of the four NN atoms per unit mesh, three also bond to Cu atoms in the layer below and are therefore five-coordinate. Two of the three five-coordinate sites result from appreciable lateral displacements toward each other of both N atoms and Cu atoms below in the topmost Cu(110) layer. Average N-Cu bond lengths for the five-coordinate sites (1.87 Å) agree well with prediction, while the bond lengths for the four-coordinate sites (1.85 Å) appear somewhat long. These different N sites are discussed in relation to a recent STM study, and results from this LEED analysis are contrasted with those from the previous LEED study, which supported a 〈001〉-missing row reconstruction of the surface.

A H-T superconducting phase diagram is presented for ${\mathrm{UPt}}_3$ which is obtained from ultrasonic velocity meausrements with the field at 45° to the c axis. The implications of the phase diagram on various theoretical models are discussed.

Dynamics of a ‘‘quasi-two-dimensional’’ flux-line lattice in the layered superconductor 2H-${\mathrm{NbSe}}_2$ for field normal to the layers is investigated at high fields. Comparisons with numerical simulations imply the presence of plastic flow for soft lattices. Dimensionality effects are more pronounced in the disorder-dominated regime, where a power-law scaling form for the I-V curve yields the conductivity exponent above the percolation threshold.

With use of numerical simulation we investigate the influence of self-fields on the static properties of one-dimensional parallel arrays of Josephson junctions. We determine the currents and fields around a vortex, and the critical current versus magnetic field when the mutual inductances between all cell pairs are included. We find that for these static properties a good approximation can be obtained by using self-inductances only, but with a larger effective penetration depth which is determined by the ratio of the nearest-neighbor inductance to the self-inductance of the loops. We also find that for small penetration depths there is an energy barrier for vortex motion from one cell to the next but inclusion of all the mutual inductances only has a minor influence on this energy barrier.

We have calculated the elementary-spin-excitation spectrum of undoped and doped Hubbard models in one and two dimensions. We have obtained the gap of the spectrum at Q=(π,π) in the doped strong-coupling Hubbard model by a Lanczos diagonalization study of the √10 × √10 site cluster. The result is qualitatively in agreement with the experimental observation of the ‘‘spin gap’’ by Rossat-Mignod et al. We conclude that the ‘‘spin gap’’ can be accounted for with the doped strong-coupling Hubbard model without any symmetry reductions and is one of the intrinsic properties of the Hubbard model.

${\mathrm{Bi}}_2$${\mathrm{Sr}}_2$${\mathrm{CaCu}}_2$${\mathrm{O}}_{\mathit{y}}$/Au point-contact junctions have been studied. For a description of junction properties we used the Blonder-Tinkham-Klapwijk theory where we included the finite lifetime of the quasiparticles. From the fitting procedures of the experimental differential characteristics, the values of the Fermi velocity ${\mathit{v}}_{\mathit{F}}$≃5×${10}^7$ cm/s, the ratio 2Δ/${\mathit{k}}_{\mathit{B}}$${\mathit{T}}_{\mathit{c}}$≃8, and the coherence length ${\xi }_{\mathit{a}\mathit{b}}$≃35 Å for ${\mathrm{Bi}}_2$${\mathrm{Sr}}_2$${\mathrm{CaCu}}_2$${\mathrm{O}}_{\mathit{y}}$ were obtained.

We consider the two-channel Kondo model in the Emery-Kivelson approach, and calculate the total susceptibility enhancement due to the impurity ${\mathrm{\chi }}_{\mathrm{imp}}$=χ-${\mathrm{\chi }}_{\mathrm{bulk}}$. We find that ${\mathrm{\chi }}_{\mathrm{imp}}$ exactly vanishes at the solvable point, in a completely analogous way to the singular part of the specific heat ${\mathit{C}}_{\mathrm{imp}}$. A perturbative calculation around the solvable point yields the generic behavior ${\mathrm{\chi }}_{\mathrm{imp}}$∼log(1/T), ${\mathit{C}}_{\mathrm{imp}}$∼T logT and the known universal value of the Wilson ratio ${\mathit{R}}_{\mathit{W}}$=8/3. From this calculation, the Kondo temperature can be identified and is found to behave as the inverse square of the perturbation parameter. The small-field, zero-temperature behavior ${\mathrm{\chi }}_{\mathrm{imp}}$∼log(1/h) is also recovered.

The magnetic properties of ultrathin (1–5 monolayer) Fe films on Ag(100) substrates were investigated using SQUID magnetometry. Films were grown in pairs (one bulklike, the other thin), and characterized in situ by low-energy electron diffraction, Auger spectroscopy, and the surface magneto-optic Kerr effect. The films were than capped with Au and studied with a SQUID magnetometer over the temperature range 2–340 K. We report here a direct observation of enhanced magnetic moments for Fe on Ag(100), with interface moments enhanced as much as 29%.

A set of integral equations in the Fermi-hypernetted-chain scheme is formulated for Gutzwiller correlated wave functions. The well-known exact analytical results of Vollhardt, Metzner, and Gebhard for the one- and infinite-dimensional Gutzwiller paramagnetic wave functions are recovered as a byproduct of the theory. In approximate form, the theory is applicable to arbitrary Gutzwiller correlated wave functions, irrespective of the space dimensionality of the model.

Using an electronic theory we present calculations of the magneto-optical Kerr rotation angle in second-harmonic generation (SHG). For the longitudinal and polar Kerr configuration and for arbitrary angles of incidence it is found that the Kerr angle in SHG may be enhanced by up to one order of magnitude compared to the linear Kerr angle. This enhancement is caused by interband and intraband transitions (plasmons), which in the linear case suppress the Kerr rotation in the optical range. Our results will be useful for a microscopic study of two-dimensional magnetism.

The magnetic properties of ${\mathrm{YBa}}_2$(${\mathrm{Cu}}_{0.96}$${\mathrm{Zn}}_{0.04}$${)}_3$${\mathrm{O}}_{\mathit{x}}$ were studied in detail by means of muon spin rotation and relaxation for 6.0≤x≤6.92. The complete magnetic phase diagram was mapped out and a disordered magnetic state was found to persist between x=6.4 and x≃6.7 (metallic transition), in contrast with pure ${\mathrm{YBa}}_2$${\mathrm{Cu}}_3$${\mathrm{O}}_{\mathit{x}}$. The appearance of this magnetic state is attributed to the effect of Zn on the doped hole dynamics and might be associated with the freezing of local moments due to Zn (6.43≤x≤6.88), which were also detected here, in the paramagnetic state.

The magnetoelectric (ME) effect is observed in the tetragonal antiferromagnet ${\mathrm{Gd}}_2$${\mathrm{CuO}}_4$ in magnetic fields up to 7 T. This is an observation of the ME effect in the class of tetragonal crystals with the general formula ${\mathit{R}}_2$${\mathrm{CuO}}_4$ (R=rare-earth), which are parent compounds for high-${\mathit{T}}_{\mathit{c}}$ superconductivity. The ME effect is shown to exist only below the ordering temperature of the rare-earth magnetic subsystem at ${\mathit{T}}_{\mathit{N}}$(Gd)=6.5 K in accordance with results based on a symmetry analysis. The field and temperature dependence of the effect confirm the magnetic structure revealed recently in this compound.

Zn-impurity-induced anisotropy reduction of superconductivity is obtained from a resistive transition analysis of single-crystal ${\mathrm{YBa}}_2$(${\mathrm{Cu}}_{1\mathrm{-}\mathit{x}}$${\mathrm{Zn}}_{\mathit{x}}$${)}_3$${\mathrm{O}}_{7\mathrm{-}\mathrm{\delta }}$. The out-of-plane resistivity (${\mathrm{\rho }}_{\mathit{c}}$) increases with Zn doping, while the carrier concentration remains almost unchanged. The results are not consistent with the simple Lawrence-Doniach model. Instead, an anisotropic three-dimensional metal model seems to be promising. A Zeeman-contribution-subtracted magnetoresistance analysis indicates that a Zn impurity does not cause any magnetic pair breaking.

The variations of voltage and voltage noise with magnetic field in c-axis-oriented ${\mathrm{DyBa}}_2$${\mathrm{Cu}}_3$${\mathrm{O}}_{7\mathrm{-}\mathrm{\delta }}$ thin films exhibit reproducible, microstructure-dependent ‘‘magnetofingerprints’’ (MF’s). The MF’s can be scrambled with the reversal of the direction of the Lorentz force on field-induced vortices. Analysis of the noise suggests a strong dependence of the local free-energy landscape on vortex density similar to global frustration effects found in periodic superconducting networks. Above fields which roughly match vortex separation with mesoscopic-scale disorder in the films the noise and the fine structure of the MF’s are suppressed, suggesting a crossover to a more weakly pinned vortex-liquid regime.

The system Ce(${\mathrm{Cu}}_{1\mathrm{-}\mathit{x}}$${\mathit{M}}_{\mathit{x}}$${)}_{2.2}$${\mathrm{Si}}_2$ was doped with M=Pt, Ir, Os, and Ni, always maintaining the original ${\mathrm{CeCu}}_2$${\mathrm{Si}}_2$ structure. The doped samples as well as the end points (i.e., x=1) are nonmagnetic. Measurements on the specific heat C of lower-doped samples (x<0.2) showed a monotonic decrease of γ with increasing doping. This decrease depends only on the amount of doping and not on M, thus not on the variation of the lattice constant or the electronic structure of the dopants. We conclude that the decrease in γ is primarily determined by the disturbed coherence in the heavy-fermion lattice, and that variation of the Ce-M ligand hybridization merely plays a secondary role. This result underlines the fact that coherence is a decisive prerequisite for the formation of a heavy-fermion state, and that this coherence is equally dependent on disturbance of the ligand sublattice as of the f-atom sites.

Measurements of the short-wavelength ${\mathrm{\Sigma }}_4$ phonons in a virgin crystal of Li metal at 293, 200, 100, and 80 K show a slight softening of the phonons with a decrease in temperature, as observed in some other alkali metals, contrary to all the other phonon branches measured in high-symmetry directions. Theoretical anharmonicity calculations predict this anomalous behavior even for Rb, which does not undergo a transformation. Diffuse scattering measurements in the vicinity of the (110) and (200) reflections at 200, 100, and 80 K did not reveal any unusual changes. There is an overall increase in the background as the temperature was decreased, which is attributed to the Debye-Waller factor associated with the large spin-incoherent scattering. The widths of the phonon groups also decreased as the temperature was lowered. These and other neutron-scattering measurements, along with ultrasonic attenuation studies fail to give evidence for precursor effects of the 9R transformation in virgin crystals of lithium metal.