Table of contents

The effect of coupled-mode LO phonons on the momentum relaxation of electrons is studied within a model that takes the internal thermalization time of the LO-phonon system and the relaxation time of the whole lattice as parameters. Our analysis shows that the LO-phonon-limited electron mobility is markedly reduced at low temperatures due to the formation of finite-lifetime coupled modes in a polar semiconductor.

Magnetization measurements in the temperature range 1.5-300 K have been made on two series of samples of nanocrystalline ferromagnets: Fe₆₀Co₃₀Zr₁₀ and Fe_73.5CuNb₃Si_13.5B₉. It was found that at low temperatures magnetization decreases in accordance with the Heisenberg-model prediction for both series of samples. The coefficients B and C were obtained by fitting M_s(T) to two terms, T^3/2 and T^5/2, from the Bloch law. It is shown that for both series of samples the B and C coefficients are larger in the amorphous state than in the nanocrystalline state; larger values of spin-wave stiffness constants D were obtained in nanocrystalline FeCoZr samples with grain size d<40 nm than in the amorphous state; with increasing grain size a much larger value for D was observed for nanocrystalline FeCoZr with a grain size of 140 nm. The results present a reasonable agreement with predictions for nanocrystalline materials with two or three different magnetic phases.

Self-consistent parameter-free calculations of the surface electronic structure of a random alloy are used to discuss the compositional dependence of the work function ( Phi ) of the PdAg(111) random alloy surface. For crystals with uniform concentration profile Phi varies almost linearly between the pure metal limits and in the presence of segregation is found to be a measure of the surface layer composition. Good agreement with measured polycrystalline work functions is found using experimentally determined segregation data. The results are used to understand screening within PdAg alloys and point to possible surface-alloy formation during the growth of Pd on Ag.

The authors have measured the Ga⁰ emission yield from GaAs(110) surfaces for laser pulses of several photon energies near the band-gap energy 1.435 eV, ranging from 1.33 eV to 2.53 eV. Similarly to the case for emissions of Si atoms from Si(100) and of Ga atoms from GaP surfaces, they find that the Ga⁰ emission yield, under repeated irradiation with laser pulses at fluences smaller than that for ablating the surface, decreases from its initial value rapidly at first and then slowly, while repeated irradiation by laser pulses above the ablation threshold fluence increases the emission yield gradually. It is found that the Ga⁰ emission yield for laser pulses below the ablation threshold depends strongly on the photon energy h nu : the emission yield is relatively small for h nu <1.39 eV (region I); the emission yield is essentially zero for 1.39 eV<h nu <1.43 eV (region II); and the yield increases in a stepwise fashion when the photon energy increases across the transition energies involving surface states for h nu >1.42 eV (region III). It is suggested that the emission for region I is induced by electronic excitation of defects on the surface, while that for region III is induced most dominantly by electronic transitions involving surface states. It is also found that the yield is a power function of the laser fluence with power indices 2-4 for the rapidly decaying component and 4-7 for the slowly decaying component, depending on the photon energy.

The authors have used a self-consistent semi-empirical molecular orbital method to investigate whether the adsorption properties of K atoms and formation of the K adsorbate chains or clusters in the low-coverage regime can be influenced by the nature of the semiconductor surface: perfect or stepped. In the process, they are able to determine the microscopic structures of monatomic and diatomic K molecules on perfect and stepped GaAs(110) surfaces. The results for K adsorption on the perfect GaAs(110) surface are consistent with recent scanning tunnelling microscopy (STM) observations for Na on GaAs(110), with the stable site for K being the bridge site encompassing one Ga and two As surface atoms. The equilibrium geometry for diatomic K has the second K atom occupying the next-nearest-neighbour bridge site, strongly supporting the information of an open linear structure parallel to the surface atomic zigzag chains. The calculated K-K distance in this equilibrium configuration is 8.02 AA, similar to the Na-Na distance (8 AA) from the STM experiment. The results for the stepped GaAs(110) surface suggest that a step is unlikely to assist clustering of K atoms, but, the formation of the linear adsorbate chain appears rather to be influenced by the orientation of the steps. However, the K adsorbates are bound more strongly at the steps than at the bridge sites on the perfect surface.

By means of an electron beam deposition technique under ultrahigh vacuum Co/Au(001) superlattices were prepared on GaAs(001) substrates with a 500 AA thick Au(001) buffer layer. The superlattices were characterized by reflection high-energy electron diffraction (RHEED), and X-ray diffraction. The authors found that well defined single-crystalline Co/Au(001) superlattices can be obtained only for samples with Co layer thickness below about 10 AA. The results from X-ray diffraction measurements indicate a lightly strained BCC structure for Co layers with the lattice constants a/sup ///=2.88+or-0.01 AA in the film plane and a^{perpendicular to} =2.78+or-0.02 AA. along the growth direction.

In the X-ray diffraction pattern from a thermally oxidized thin film on an Si(001) surface, very weak Bragg peaks have been observed. The thermally oxidized thin film is, therefore, not purely amorphous but many small crystallites are dispersed within it, maintaining an epitaxial relation with the Si substrate. It is difficult to determine the structure of the crystalline phase because of the limited number of observable Bragg peaks. The pseudo-cristobalite structure proposed by Iida et al. was selected as a possible model. The atomic arrangement is similar to that of the cristobalite structure, while the unit cell is tetragonal so as to match the lattice spacing to that of the Si substrate. A least-squares fitting analysis of the profile of the newly observed Bragg peak reveals that the crystallites are located not only at the interface between the Si substrate and the amorphous layer, but also widely distributed in the amorphous layer, preserving an epitaxial relation among the crystallites. The proportion of such crystallites is estimated to be a few percent of the whole volume of the amorphous layer.

The crystallization of Co and Au(111) has been shown to be possible on molybdenite substrates at 650 K as proved by good LEED patterns. The growth is, however, affected by islanding, UPS evidences a structure attributed to the Co and Au. The same kind of measurement has been performed for Au/Co/Au and Co/Au/Co/Au sandwiches with identical issues.

The interaction of a sodium atom with the silicon(111) surface is studied from the point of view of molecular orbital theory. The Si surface is modelled by finite clusters and the interaction of the alkali atom is investigated from different approach positions. Electron correlation is incorporated by invoking the concepts of fourth-order many-body perturbation theory. The results are compared with available data in the literature.

The authors have carried out a theoretical investigation of the surface electronic structure of the relaxed GaP(110) surface and of a monolayer of Sb on GaP(110) by a self-consistent tight-binding method. These calculations, yielding band structure and local densities of states, show some marked differences from other calculations. In addition they have been able to determine the amount of charge transferred between the Sb overlayer and the substrate.

The authors consider the electronic transport process in thin films with rough surfaces. The approach employed is based on classical concepts and avoids some special assumptions used in earlier papers. Roughness is described in terms of scatterers that are distributed randomly on the surfaces. Their differential scattering cross section yields a generally angle-dependent specularity parameter. A new form of this angle dependence is proposed. They show that the homogeneity of the lateral carrier distribution, as postulated by Fuchs in his well known paper published in 1938, is associated with the time-reversal symmetry of the microscopic scattering cross section. The divergence of the classical surface-dominated conductivity, i.e. for vanishing bulk scattering, is enhanced for angle-dependent specularity parameters in comparison with the constant parameter introduced by Fuchs. They further show how to make contact between classical formulations and quantum-mechanical approaches leading to a stronger dependence of the conductivity on the film thickness. The results are compared with existing classical size-effect theories.

The author studies the plasma oscillations in a finite number of layers of a 2DEG. It is shown that by matching electromagnetic boundary conditions and solving the recurrence relations for the coefficients of the electric potential in different regions, a general equation can be obtained for the frequencies of all the plasmons under different types of surface discontinuity. Numerical results are shown for six layers with nonidentical surface charge layers.

The authors study the electronic structure of the epitaxially grown Pd monolayer on the Ta(011) surface as well as the electronic properties of other related systems (Pt overlayer; Nb, Mo and W substrate). To this end an s-d electronic tight-binding Hamiltonian is treated using the recursion method technique. The calculations confirm that the 'noble-metal' properties of the overlayer are mainly due to the overlayer-substrate electron hybridization in which the role of s electrons is not negligible. The possible contribution of initial-stage effects to core-level shifts is also considered. For Pd adatoms they obtain a semi-quantitative agreement with experiment.

Presents a theory of electromagnetic (EM) wave propagation through magnetic multilayers and superlattices based on the propagation matrix P of a magnetic film. By using this matrix P, the authors obtained the transmission and reflection coefficients of layered magnetic media, including semi-infinite magnetic surfaces, magnetic multilayers and semi-infinite magnetic superlattices. The numerical results show that the EM modes of a magnetic layer system are excited and manifested as the sharp dips in the s-polarized reflection and the dispersion curves of magnetic polaritons can be measured by a method similar to the attenuated total reflection technique. The analytic dispersion relations of magnetic polaritons for arbitrary magnetic multilayer systems are represented by the matrix P.

Interface plasmon excitations of superlattices with defects are investigated by the propagation matrix method. A dispersion relation is obtained and is shown to be of sufficient generality. The dispersion curves of the local modes are shown to be dependent upon the thickness and the dielectric constant of the defect layer. The obtained formula generalizes some earlier results obtained by other authors. Several special cases, including the limiting cases of quantum-well layered structures, are discussed and a new mode similar to the Giuliani and Quinn type mode is deduced.

Chlorinated Si(111) 7*7 surfaces were examined with scanning tunnelling microscope images after pulsed laser irradiation at 266 and 355 nm with low laser fluence, where the thermal effect can be ignored. From the surface irradiated by the 266 nm laser, dichloride and trichloride species are desorbed, while monochloride species remain on the surface. This desorption selectivity was not observed at 355 nm. These results give useful information to elucidate the formation mechanism of a stripe pattern observed after 266 nm laser irradiation on a Cl-saturated Si(111) 7*7 surface.

The interaction of atomic hydrogen with the cleaved GaAs(110) surface has been investigated by high-resolution electron energy loss spectroscopy (HREELS), at primary energies of 5 and 15 eV, analysing the losses associated with three different mechanisms: (i) low-energy surface collective excitations (surface TO phonon and dopant-derived free-carrier plasmon); (ii) the stretching of the Ga-H and As-H bonds: and (iii) the electronic losses above the fundamental gap, involving both surface and bulk electronic states. The first kind of loss is very sensitive to H exposure and shows that hydrogen induces a band bending at the lowest exposures. The vibrational part of the spectrum indicates that the exposure of 10⁴ L corresponds to a coverage of one monolayer and that H bonds to both Ga and As over the whole coverage range. The region of the electronic transitions indicates the disappearance of transitions of the clean surface and the appearance of new transitions characteristic of the H covered surface. At high exposures the growth of a very strong background is consistent with the presence of small metallic Ga clusters. This result is also consistent with the modifications, at the same exposures, of the low-energy region of the spectrum.

Several homogeneous, ordered or disordered, Si(111) surfaces namely 1*1-H, 7*7-H, square root 3-Ga, square root 3-Ag and oxygen disordered, have been prepared in ultra-high vacuum and their photoemission yield spectra have been measured in the threshold photon energy range (4-7 eV). As expected along surfaces where the surface state contribution is very small, the spectra obey a power law of the form (h nu -E_i)^alpha , E_i being the ionization energy; alpha is found to depend strongly on the surface geometry, being two on nearly non-scattering surfaces (the hydrogenated ones) and reaching ⁷/₂ in a highly scattering case (oxygenated surface) in agreement with theoretical predictions. This leads to an accurate determination of E_i. The validity of this treatment is discussed for surfaces in which the surface state contribution is important.

The authors propose a simple model for (CO)_1-x(N₂)_x mixtures adsorbed on graphite where a lattice site carries a spin S_i=+or-1 representing the orientations of the CO electric dipole moment if the site is occupied by a CO molecule, while the spin S_i=0 if the site is taken by an N₂ molecule, which has a quadrupole moment only. Assuming a bilinear dipole-quadrupole coupling, randomly quenched N₂ impurities then act as random fields would act on an Ising antiferromagnet. For simplicity, a square lattice is treated and the range of all interactions is restricted to nearest neighbours. Monte Carlo studies are performed for a range of lattice linear dimensions L from L=24 to L=50, and the specific heat order parameter and susceptibility, as well as the fourth-order cumulant, are studied, applying finite-size scaling concepts where appropriate. The specific heat results display a striking qualitative similarity to the experimental data of Wiechert and Arlt, and provide evidence that the transition is already rounded by arbitrarily small dilution, consistent with theoretical predictions for the two-dimensional random field Ising model. While the experiments needed to rely on the specific heat only, the data for the (strongly rounded) ordering susceptibility and the cumulant (where the common intersection point disappears, consistent with the absence of a transition) provide compelling evidence for this picture. The crossover scaling analysis first proposed by Ferreira et al. for dilute antiferromagnets in a field also works out reasonably well for the present model.

The authors report measurements of carrier mobilities in thin transparent films of amorphous indium oxide using the field effect. The field-effect mobilities (of order 10^-3 m² V^-1 s^-1) are similar to those calculated from conductivity and Hall effect measurements. Their similarity sets an upper limit of order 3 * 10¹⁶ m^-2 to the density of surface-trapped charge. The mobilities are temperature independent, consistent with being determined by ionized donor scattering which is known to dominate in these systems. They show that conduction can be turned on in a device prepared on the insulator side of the metal-insulator transition with mobilities near threshold about an order of magnitude smaller than those observed at higher carrier densities. The difference is attributed to reduced screening. Failure to invert the channel suggests a significant density of traps in addition to the oxygen vacancy donors.

A general method is suggested to determine the site preference of alloying additions in intermetallic compounds with either stoichiometric or non-stoichiometric compositions, by first-principles binding-energy calculations using cluster models. The relationships of the site preference of alloying additions, with the compositions of the host elements and with the concentrations of the additions themselves in the host are established. The validity and reliability fo the method are tested by applying it to the site preference study of some ternary additions (Sc, Ti, V, Fe, Co, Cu, Nb, Mo and Pd) in gamma '-Ni₃Al. The results obtained are in complete agreement with the experimental data, which proves that the method is successful.

Probability distributions of fluctuations of atomic positions around scattering vector dependent reference lattice point have been calculated for various crystalline and quasicrystalline structures and their influence on the diffraction pattern has been discussed. For incommensurate scattering vectors flat fluctuation distributions are found, which change drastically when the scattering vector approaches the diffraction peak position. Moments of these distributions have been used to calculate the diffraction pattern. Limitations of the Debye-Waller approximation are discussed and a new more accurate method of peak intensity estimation is proposed and tested for real space and phason space fluctuations.

Ag_xTiS₂ intercalation compounds were thermally prepared and characterized by a room-temperature X-ray diffraction method to be of single phases in stage 1, stage 2 and stage 1' in the ranges x=0.36-0.42, 0.17-0.22 and 0-0.12, respectively. The heat capacities of Ag_0.361TiS₂ were measured by adiabatic calorimetry in the temperature range between 13 and 400 K. The intralayer order-disorder phase transition of silver ions was observed at 285.5+or-0.3 K as a higher-order type with wide skirts of anomalous heat capacity on the low- and high-temperature sides of the peak. The enthalpy and entropy of the transition were found to be 4.3+or-0.4 kJ mol^-1 and 9.9+or-1.0 J K^-1 mol^-1 (close to R In 3), respectively, on the assumption of a high-temperature approximation. The interaction of positional ordering of silver ions is suggested to be of a two-dimensional nature, and two theoretical models relevant to the phase transition are discussed.

The electrostatic potential in an ionic crystal is presented in terms of the zeroth-order MacDonald function by means of decomposition of the crystal lattice into sets of parallel lines with the periodic charge distribution with the zero net charge within a period. The method is applied to the calculation of the Madelung constants of a series of standard cubic lattices, doped fullerides and idealized Y-Ba-Cu-O crystals.

The authors present results from electronic structure calculations of the intermetallic hydrides M₂M'H₆ (M=Mg, Ca, Sr, M'= Fe, Ru, Os). Most of these hydrides have been recently synthesized and characterized, all showing the K₂PtCl₆ cubic structure. They have determined their energy bands and densities of electronic states by means of the non self-consistent augmented-plane-wave method. The results show that the nine hydrides are semiconductors with relatively large energy gaps. These isostructural and isoelectronic families of hydrides have been studied in order to establish correlations between the ordering of the electronic states and the energy gaps in terms of differences in metal-hydrogen distances and atomic energy level. They observe that the splitting between the bonding and antibonding M' d(e_g) orbitals is responsible for the increase of the energy gap as the atomic number of M'increases. Nevertheless, the presence of d states from the divalent metals Ca and Sr modifies the nature of the unoccupied conduction states above the energy gaps. In general, a reduction of the energy gap is obtained in the Ca- and Sr-derived hydrides.

Scattering rates of Auger and electron-phonon processes between discrete Landau levels in a quasi-two-dimensional electron system are calculated. Using these rates the authors provide an understanding of the upconversion observed in the magneto-luminescence of one-side modulation doped GaAs/AlGaAs quantum wells; in particular, of its dependence on excitation power and magnetic field.

The ground state energy of large polarons is investigated by means of squeezed states. First the authors introduced single-mode squeezed states and used the variational method to calculate this energy. This gives a result that is valid in the weak-coupling regime and only for very small values of the coupling constant. In order to improve the result they have considered two-mode squeezed states in which the correlation between phonons is involved. In this way the ground state energy is found to be lower than the Feynman result in the intermediate-coupling regime for alpha < 3.7.

The authors report the electrical resistivity rho , thermopower S and thermal conductivity lambda studies of (Sm_1-xY_x)Cu₂ and RCu₂ (R identical to Gd, Pr or Tb) systems between 4.2 and 300 K. The value of the antiferromagnetic ordering temperature T_N for the ((Sm_1-xY_x)Cu₂ systems is found to become depressed, almost linearly, with Y substitution. The value of T_N extrapolates to zero as x to 1. Such a behaviour is attributed solely to the dilution effects of Sm by Y atoms. Chemical pressure effects on the value of T_N due to addition of Y atoms are found to be negligible. The lambda curves of SmCu₂, GdCu₂ and TbCu₂ exhibit a peak for T < T_N. This low-temperature peak at T_max^lambda approximately= 10 K, in the lambda data of these compounds arises from the combined influence of the electronic and the magnetic contribution to the total lambda . This low-temperature peak becomes suppressed for x >or= 0.10. The S curves exhibit several extremum features, at T <or= T_N, which is ascribed to the variation in the density of states at the Fermi level.

Grain-type alloy thin films consisting of fine ferromagnetic grains, which are composed of Fe, Co, Ni, or their alloys, in non-magnetic Ag, Bi, or Mg matrices were prepared by a vacuum deposition method. Structural, magnetic, and magnetoresistance characteristics were studied on such films as a function of the composition and concentration. A giant magnetoresistance GMR effect was observed over a broad range of magnetic concentrations in the Ag-based systems, except the Ni-Ag one, at room temperature. The appearance of the GMR effect is probably correlated with the Curie temperature of the magnetic component and the atomic radius ratio of the magnetic and non-magnetic components.

Microwave absorption was measured in highly oriented ceramic samples of YBa₂Cu₃O_6.9 obtained by melt-quenching-melt growth solidification. Measurements on samples cooled down to the superconducting state in an external magnetic field permitted analysis of magnetic flux trapping. The character of the magnetic flux trapping has been found to be isotropic despite distinct structural anisotropy of the studied materials confirmed by the results of X-ray analysis. An attempt to interpret the results within the model of intergrain Josephson junctions has been made.

The magnetic properties of BaBiO₃ and the metal-substituted compounds BaBi_0.5In_0.5O₃, BaBi_0.85Mo_0.15O₃ and BaBi_0.5La_0.5O₃ were measured. A combined behaviour of temperature-dependent Curie-like paramagnetism and of independent diamagnetism was observed. The diamagnetic contribution is about 8 * 10^-5 emu mol^-1 almost independently of the metal impurities. The paramagnetic contribution, on the other hand, shows markedly different behaviours for different impurities. The number of Bohr magnetons included in a molecule is larger by one order of magnitude for the compound with indium than for those without and with transition-metal impurities. It was concluded that the paramagnetic behaviour comes from the holes on the Bi 6s and O 2p bonding orbits, while the diamagnetic behaviour mainly comes from the core electrons on barium or on the other ions. A deviation from the Curie law in the low-temperature region was observed. In the case of BaBi_0.85Mo_0.15O₃ and BaBi_0.5La_0.5O₃ the behaviour could not be explained by the simple idea of saturation in a field. For BaBi_0.5In_0.5O₃ a phase-transition-like behaviour at around 10 K was observed. It is pointed out that there is some correlation between the number of Bohr magnetons and the number of electric current carriers.

Inelastic neutron scattering experiments investigating the spin dynamics along the chain direction of an S = ³/₂ quasi-one-dimensional Heisenberg antiferromagnet CsVCl₃ were performed up to the zone boundary energy. The authors observed the quantum renormalization in the exchange constant and found that the scattering function cannot be described by either the classical delta -function or the quantum Heisenberg model for S = ¹/₂. The spin excitations have apparent linewidths and they are compared with the theory of the classical isotropic Heisenberg system at finite temperatures.

The two-photon odd-parity vibronic transitions in centrosymmetric systems have been described by a simple model based on the third-order mechanisms. The use of symmetry-adaptation techniques and Racah-Wigner calculus has led to the expression for transition intensity, including magnetic and vibronic components as well as the polarization factor. This expression has been used for calculations of intensity strengths of two-photon vibronic transitions between the ground state ((⁸S₇2/) mod and one of the excited J multiplets >>(⁶P₅2/)) of the Gd³⁺ ion in surroundings of octahedral symmetry.

Antiphase boundaries play an important role when considering the mechanisms involved in the plastic deformation of a face-centred cubic alloy. The authors consider (100) antiphase boundaries in a binary L1₂ ordered alloy to which a small amount of ternary element is added. The scope of this work is to study the behaviour of the ternary element at the antiphase boundary (segregation, effect on the energy of the antiphase boundary) and to determine general simple rules that can help to predict this behaviour.