We introduce a method to study the activation barrier to nucleation of dislocations during strain relaxation of thin films. By growing on substrates with a range of misorientations from (001), we change the applied shear stress on the different glide systems, leading to an imbalance between nucleation on the different glide systems. The resulting tilt between the relaxed layer and the substrate provides a direct measure of the stress derivative of the nucleation barrier. A phenomenological model, which makes no assumptions about the detailed nucleation mechanism, can then be used to infer from this the activation energy for nucleation. We obtain an activation energy of 5 eV for the observed nucleation mechanism.

Time-resolved reflection high-energy electron diffraction is used to study the structural behavior of Pb(100) subjected to 200 ps laser pulses. Analysis of the electron diffraction intensity indicates that residual order is present on the surface at a temperature of 615±20 K, where the bulk melting temperature of Pb is 600.7 K. This implies that the disordered film that forms on Pb(100) is stable up to the bulk melting temperature and above.

To determine the behavior of the line-point tension τ^ at the wetting transition, we consider a two-dimensional Ising model with appropriate boundary conditions and investigate suitable definitions of τ^ associated with the junction of an interface tilted with average angle ${\mathrm{\theta }}_{\mathit{c}}$ and another lying along the substrate. Size-dependent fluctuations in the point of contact require that τ^ be defined through a convolution sum. Hence τ^≊ln(1/${\mathrm{\theta }}_{\mathit{c}}$) as ${\mathrm{\theta }}_{\mathit{c}}$→0 (wetting transition), which can be understood as a consequence of the entropic repulsion of the tilted part of the interface against the substrate.

It is shown that the freedom in the choice of the proposal matrix in the generalized Metropolis method can be used to greatly enhance the efficiency of the method. For example, the difficulties associated with the existence of multiple length scales in electronic structure calculations can be avoided by making an intelligent choice. Results are presented for Ne, Ar, and ${\mathrm{Li}}_2$.

We study analytically statistics of a quantity known as an inverse participation ratio P which is inversely proportional to a spatial extent of localized eigenfunctions. The fluctuations are found to be crucially dependent on the ratio between the system size and mean localization length. As a particular model, we use an ensemble of random banded matrices which is an equivalent way to describe wires with a large number of transverse modes. Our results are in agreement with available numerical data for periodically driven Hamiltonian systems in the quantum chaos regime.

From a detailed study of two different defects in silicon, excitonic Auger capture is shown to be an important mechanism of the configurational change of metastable defects. A free exciton interacts with the defect and recombines, transferring its energy to an electron, which will be emitted high up into the conduction band. This hot Auger electron will quickly reach the bottom of the conduction band. The liberated energy may excite the defect above the barrier between the two configurations.

The LO phonon resonant Raman scattering is studied in GaAs/${\mathrm{Al}}_{0.32}$${\mathrm{Ga}}_{0.68}$As Bragg confining structures and in a similar (reference) superlattice. Strong resonances, with a large outgoing/incoming beam intensity ratio, are observed in the spectral range of the (${\mathit{e}}_{\mathit{B}}$:${\mathit{hh}}_{\mathit{B}}$) Bragg confined excitons as well as in the (${\mathit{e}}_1$:${\mathit{hh}}_1$) exciton band. The resonance profiles are analyzed in terms of a model of exciton scattering by interface and alloy potential fluctuations. The Bragg confined 1S excitons are found to be virtually two dimensional, while those of the superlattice are intermediate between two and three dimensions.

The electron spectral function of a 2D interacting electron system in a strong magnetic field is found to exhibit an energy gap at arbitrary filling fractions. In general, this gap is larger than the energy cost associated with a charge e excitation. At this ν=1/2 filling fraction, a numerical estimate of the tunneling gap is ${\mathrm{\Delta }}_{\mathrm{tun}}$=(3.0±0.1)${\mathit{e}}^2$/ε${\mathit{l}}_{\mathit{B}}$ for the Coulomb interaction.

Using a novel fabrication technique, we have studied the metal-insulator transition in in situ grown wires of Pb with cross-sectional areas as small as ${10}^{\mathrm{-}13}$ ${\mathrm{cm}}^3$. We find that as we cross over into the 1D regime, fluctuations dominate the superconducting transition. Although existing fluctuation models work well at the larger cross-sectional areas, there is systematic deviation as the wires become smaller. We speculate that this may be due to 1D Coulomb correlation effects.

We report on a Monte Carlo study of two-dimensional Ginzburg-Landau superconductors in a magnetic field. We find evidence suggestive of a first-order phase transition between a low-temperature state characterized by power-law positional correlations in the superfluid density and a high-temperature vortex-fluid state. A key aspect of our study is the introduction of a quantity proportional to the Fourier transform of the superfluid density which can be sampled efficiently in Landau gauge Monte Carlo simulations and which satisfies a useful sum rule.

We report on thermal cycling of an ${\mathrm{YBa}}_2$${\mathrm{Cu}}_3$${\mathrm{O}}_7$ single crystal close to the melting temperature of the flux lattice for H‖‖ab. Transport current is along the c axis and the temperature stability is better than 1 mK. The melting (freezing) transition of the flux lattice occurs at different temperatures (ΔT=7 mK) between the cooling and the heating cycle. Hysteresis disappears for slightly increased current density (0.1 A/${\mathrm{cm}}^2$). Comparison of thermal cycles reveals that the Abrikosiv lattice superheats, in contrast with the usual behavior in real crystals where surface premelting is important. This can be due to its imbedding in the real crystal and the consequent surface pinning.

We employ ${}_{}{}^{139}{}_{}{}^{}\mathrm{La}$ NMR spectroscopy as a structural probe of metallic ${\mathrm{La}}_2$${\mathrm{CuO}}_{4+\mathrm{\delta }}$. We find a distribution of lateral displacements of the oxygen atoms which form the apices of the oxygen octahedra. We directly observe a wide distribution of structures of the La-O layer; this distribution is strongly temperature dependent. We present copper NQR spectra which reveal a second, unexplained, copper site in the ${\mathrm{CuO}}_2$ planes. We show that these features are not manifestations of disturbances of the lattice by interstitial dopants, but are the intrinsic response of ${\mathrm{La}}_2$${\mathrm{CuO}}_{4+\mathrm{\delta }}$ to doped holes.

The surface magnetism of Gd(0001) films is studied using spin resolved electron spectroscopies. Both a canted surface magnetization vector out of the plane of the film and an enhanced surface magnetic ordering temperature 60±2 K higher than the bulk Curie temperature are observed. Our results show that the surface magnetic layer behaves as a separate entity from the bulk. The in-plane component of surface magnetization aligns ferromagnetically to the bulk. Spin resolved photoemission from the Gd 4f core levels shows ≊65% spin polarization at 150 K, which extrapolates to ≊91% at 0 K.

The satellite structure for the B Kα emission spectrum of boron oxide has been elucidated by threshold experiments carried out by varying the excitation energy of the incident undulator beam from below the B K absorption edge to several tens of eV above the absorption edge. Priminent emission of the satellite peak was observed when the excitation energy was tuned to the narrow pre-edge peak in the absorption spectrum, corresponding to the transition from B (1s) to an unoccupied antibonding p${\mathrm{\pi }}^{\mathrm{*}}$ orbital. It is concluded that the satellite peak originates from resonant x-ray Raman scattering due to a B(1s)-p${\mathrm{\pi }}^{\mathrm{*}}$-B(1${\mathit{s}}^{\mathrm{-}1}$) transition.

The shape of vesicles with genus g=2, i.e., with two holes or two handles, is studied in the framework of curvature models. These vesicles exhibit a new phase which also persists for higher genus g>2. In this phase, the ground state of the vesicle is conformally degenerate even when the volume, the area, and the total mean curvature of the vesicle are kept constant. It is predicted that such vesicles undergo a new type of diffusive motion, termed conformal diffusion, which should be observable in experiments as pronounced shape fluctuations.

Amplification studies of a two-cavity second-harmonic gyroklystron are reported. A magnetron injection gun produces a 440 kV, 200–245 A, 1 μs beam with an average perpendicular-to-parallel velocity ratio slightly less than 1. The ${\mathrm{TE}}_{011}$ input cavity is driven near 9.88 GHz and the ${\mathrm{TE}}_{021}$ output cavity resonates near 19.76 GHz. Peak powers exceeding 21 MW are achieved with an efficiency near 21% and a large signal gain above 25 dB. This performance represents the current state of the art for gyroklystrons in terms of the peak power normalized to the output wavelength squared.

We present a new Monte Carlo sampling algorithm, with which one can obtain any desired distribution of the sampling in one Monte Carlo simulation. The free energy and the entropy of a system can thus be obtained from a simple exercise of this algorithm. The main idea is to sample directly the entropy of a system at infinite temperature. Importance sampling is shown to be a particular case of the new algorithm. The algorithm is tested against the exact partition function of the L=4 simple cubic Ising model. A comparison with the multicanonical ensemble for the L=12, q=10 Potts model shows that the new algorithm is more general and more efficient.

Friedman and Sorkin have observed that the total angular momentum of an asymptotically flat, vacuum, quantum gravitational field in 3+1 dimensions need not be integral. We pursue this idea in the context of asymptotically flat 2+1 gravity, which is an exactly solvable model. We find that, for nontrivial spatial topologies, the quantized pure gravitational field has states of fractional spin. These states are dynamically allowed, in the sense that they solve all the constraints of the theory.

We show how observations of the density perturbation (scalar) spectrum and the gravitational wave (tensor) spectrum allow a reconstruction of the potential responsible for cosmological inflation. A complete functional reconstruction or a perturbative approximation about a single scale are possible; the suitability of each approach depends on the data available. Consistency equations between the scalar and tensor spectra are derived, which provide a powerful signal of inflation.

We show that a large anomalous contribution to the axial charge of the proton due to nonperturbative instantonlike gluonic field configurations implies high-energy spin effects in p-p elastic scattering. This is the same mechanism which is responsible for anomalous baryon number violation at high energy in the standard model. We compute the proton polarization due to these effects and we show that it is proportional to the center-of-mass scattering angle with a universal (energy independent) slope of order unity.