Table of contents

Tables are presented of the compatibility relations between irreducible representations of all two-dimensional space groups.

The influence of a laser field on the scattering rate for the electron-phonon interaction in semiconductor layered heterojunction structures is discussed. It is found that for a laser beam polarised in the layer plane and with a frequency omega such that omega tau >1, where tau is the carrier relaxation time, the phonon-limited mobility increases with increasing laser field strength.

Calculations of overlap integrals between conduction band and heavy hole band states with wavevectors in the (001) direction are presented for GaAs and InP. The overlap integrals are calculated using a 15-band full zone empirical k.p method, and are found to be two or three orders of magnitude smaller than conventional estimates based on effective mass sum rules. This discrepancy is resolved by evaluating all the terms that appear in such sum rules. It is found that the usual assumptions as to the dominant terms are incorrect. The small calculated values of the overlap integrals for the conduction and heavy hole bands are to be expected on group theoretical grounds.

It is shown that Laughlin's wavefunction for the fractional quantised Hall effect is not the ground state of the two-dimensional electron gas system and that its projection onto the ground state of the system with 10¹¹ electrons is expected to be very small.

The amplitude for elastic scattering of electrons by an asymmetrical two-level system (TLS) is shown to be proportional to the average TLS dipole moment, which increases on cooling. It is suggested that this may account for the increase in electrical resistance observed in metallic glasses at low temperatures.

The space distribution of the Hall current in the two-dimensional system is investigated using the one-electron form of the Kubo linear response theory. The calculation confirms that the quantised Hall current is localised only at edges of the sample when the Fermi energy lies in a gap between Landau levels of the bulk material. Unlike the bulk levels, the edge states are never separated from the Fermi energy by a gap. Therefore, the edge electrons can always respond to the applied external field.

The shape and the intensity of gamma -ray Bragg peaks in the vicinity of the ferroelastic-ferroelectric phase transition in RbH₂PO₄ and KH₂PO₄ are discussed. They reveal a distribution of the transition temperature and of the stress conjugate to the order parameter within the sample. The magnitudes of these distributions are evaluated. The temperature dependence of the diffracted intensity in the paraelectric phase is related to a severe extinction problem. The present study shows that this effect, which is often considered as an inconvenience in diffraction studies, can be favourable in the case of a phase transition. The quantitative estimates are consistent with a thermodynamic Landau-type model of the transition and how the ability of the extinction model over a large range of crystal perfection.

The authors have calculated the small-k behaviour of the structure factor S(k) of a model fluid whose interatomic potential consists of a hard-sphere repulsive core and Lennard-Jones attractive tail using the random-phase approximation (RPA) and the mean-density approximation (MDA). The attempt to assess these theories in comparison with Monte Carlo (MC) simulation results available in the literature highlighted the unreliability of the MC results in the particular region of S(k) they have studied, which they attribute to shortcomings in the algorithm used to extrapolate the radial distribution function g(r) beyond the range obtained by the computer simulation result. Moreover they argue that any attempt to develop a reliable extrapolation procedure for g(r) is inextricably linked to the availability of a good theory for the small-k behaviour of S(k); such a theory does exist at the moment. Any such theory should, in the light of this work, satisfy-as a necessary though not sufficient condition-the requirements of thermodynamic consistency.

The three disordered phases THT (tetragonal high temperature), ORT (orthorhombic room temperature) and TLT (tetragonal low temperature) of (CD₃NH₃)₂MnCl₄ have been studied by quasi-elastic neutron scattering in order to follow the reorientational dynamics of the cation CD₃NH₃⁺. Structural information has been obtained on the geometry of the motions from the incoherent structure factor measured up to Q=2.5 AA^-1 for two orientations of the crystal. In particular, it has been confirmed that the cation keeps an orthorhombic configuration in the TLT phase. Furthermore, the more probable reorientational process in all three phases occurs by 90 degrees jumps about the c axis. The corresponding rate constant decreases from about 7.1*10¹¹ jumps s^-1 at 399K (THT) to 3*10¹¹ jumps s^-1 at room temperature (ORT) and to 1.8*10¹¹ jumps s^-1 at 245K (TLT). The other kinds of reorientational motions are slower by a factor of the order of ten. From these data a direct evaluation of the probabilities of the fourth orthorhombic configurations of the cation is possible.

For pt.III see ibid., vol.17, p.217 (1984). Phase transitions of (CH₃NH₃)₂CdCl₄ (MACC) have been investigated by means of diffuse X-ray scattering experiments. The diffuse scattering patterns observed in the disordered orthorhombic room temperature (ORT) and tetragonal low-temperature (TLT) phases have been interpreted through the hypothesis that disorder involves not only the methylammonium groups but also the CdCl₆ octahedra perovskite-type layers. A simple model of planar disorder qualitatively accounts for the observed diffuse patterns. The data indicate that each layer can be considered as a mosaic of ordered 'clusters' involving about a hundred octahedra, and that disorder occurs between layers along the direction c perpendicular to the layer planes. On going into the monoclinic low-temperature (MLT) phase, the diffuse scattering disappears as expected for an ordered phase.

The transition between the incommensurate plane-wave modulation regime and the multisoliton lattice modulation regime is studied by ⁸⁷Rb quadrupole perturbed nuclear magnetic resonance in Rb₂ZnCl₄. Both phase and amplitude solitons are taken into account in comparing the experimental and theoretical lineshapes and determining the soliton density. Commensurate lines are clearly seen already in the middle of the incommensurate phase demonstrating that multisoliton lattice effects are important over a much wider temperature range than reported previously.

A general orthonormalisation procedure preserving the symmetry of the atomic orbitals is derived using the methods of Lowdin (1950), des Cloizeaux (1964) and Kohn (1973) as the starting point.

NbTe₄ and TaTe₄ form as tetragonal crystals in which the metal atoms are arranged in linear chains along the c axis, separated by Te₂ pairs. Electronic structure calculations reveal that interchain interactions are much stronger than in the related halogenated compounds of S and Se. Mixing of the Te₂p sigma * orbitals with metal d_xy orbitals induces a deep electron pocket in the lowest conduction band at the (110) zone corner, stabilising the structure against cell-doubling into a metal-metal paired semiconducting state.

GaAs grown under Ga-rich conditions often contains two dominant deep levels, sometimes referred to as the A and B levels. The authors have investigated the electronic properties of these levels in n-type and p-type material. The apparent activation energies for thermal hole emission were measured over eight orders of magnitude of thermal emission rate, and found to be 0.40 eV for level A and 0.70 eV for level B after the T² correction. All four optical cross sections have been determined. The thresholds for electron excitation are 1.15 and 0.90 eV for the A and B levels, respectively, at 80K. A photocapacitance method is used to demonstrate that the two levels are coupled and thus due to one single defect. The electronic properties, in combination with the conditions under which the defect is formed, lead to the suggestion that the defect may be the Ga_As antisite.

It is shown that the edges of a quasi-two-dimensional electron gas yield a considerable influence on the shape of the magneto-quantum oscillations of the channel resistance. In the Hall geometry, necessary for the investigation of the Hall effect, both types of front and lateral edges are of different importance for the formation of sigma _xx (or rho _xx) gaps and sigma _xy (or rho _xy) plateaus in sufficiently high magnetic fields and for the deduction of a high precision value for e²/h. In the first case, the gate-overlapped source capacitor (e.g. of a silicon MOS device) determines the equipotentials of the gate and the silicon surface allowing the chemical potential to occupy values between the broadened quantum levels of the channel. The latter effect, yielding a possibility for an accurate determination of the Sommerfeld fine-structure constant, indicates that, owing to the self-energy of the electron gas, scattering of the electrons at the lateral channel edges is completely impeded at low temperatures. Both ideas should be valid for arbitrary quasi-two-dimensional systems because of the universality of the arguments discussed.

In a recent paper the authors studied the existence of stable spin-glass phases in bond-diluted frustrated systems for various space dimensionalities and found that for d=3 a stable spin-glass phase can exist only at zero temperature. The purpose of this paper is to show that this result does not depend upon the particular real-space renormalisation technique they used.

A Migdal-Kadanoff real-space renormalisation group procedure is used to treat the critical properties of the quenched bond-mixed spin-¹/₂ Heisenberg ferromagnet on the simple cubic lattice. Within a very simple framework, the authors obtain quite reliable results for the critical temperatures. In addition, a general method for renormalising arbitrary clusters in Heisenberg spin-¹/₂ magnets is outlined.

Dielectric relaxation measurements are reported in KTaO₃:Li. Two distinct freezing processes for Li dipoles are identified: freezing of randomly interacting dipoles and locking of polar clusters.

Ionic and electronic polarisabilities in tungsten trioxide are estimated theoretically. The electronic polarisability of the O^2- ion is found to be about 10*10^-3 nm³. This is very large compared with the electronic polarisation of oxygen in other related compounds and is found to contribute about 87% to the total polarisation of tungsten trioxide. The correlation between the electronic polarisability of a W⁶⁺ ion in the free state and polarisabilities of ions in tungsten trioxide is investigated. It is found that changes in free-ion W⁶⁺ values do not significantly alter the polarisabilities of ions in the crystal. The way in which the polarisabilities vary as the dielectric constant varies between limits of 1200 and 12 in crystals containing populations of oxygen vacancies is also investigated and found not to be of significance. The possible connection between polarisabilities and defect structure formation in non-stoichiometric tungsten trioxide is discussed.

The near-band-edge photoluminescence at 80K of heavily tellurium-doped degenerate liquid-phase-epitaxial gallium arsenide layers (n=2*10¹⁸ cm^-3) is investigated. The layers are compensated with shallow germanium acceptors (the values of the degree of compensation are K=0.05-0.7) during the epitaxial growth process. It is established that even at very low concentration of shallow acceptors (K=0.05), the so-called mobile band appears in the near-band-edge luminescence spectra. The peak energy of this band shifts to lower energies, the non-equilibrium carrier generation rate decreases and saturates at values E₀^T, which are considerably lower than those calculated for the filling of the valence band density-of-states tails. It is assumed that this discrepancy between theory and experiment is due to the uncertain knowledge of the pre-exponential term in the expression for the density-of-states tails and to the formation of some deeper fluctuation states when there is a considerable difference between the concentrations of shallow donors and acceptors. It is assumed that in this case some of these states act as associates.

The core levels and valence bands of Y, YH_2.1, and YH₃ have been studied using X-ray photoemission spectroscopy. For YH_2.1 and YH₃, hydrogen-induced states appear at binding energies of about 6 eV. YH_2.1 shows emission at the Fermi level and the core-level spectra show highly asymmetric lineshape with shoulders at the high-binding-energy side of the main peaks. For YH₃, emission at the Fermi level is weak and the core-levels are nearly symmetric, consistent with a semiconductor or a semimetal. The main peaks of the core levels shift to higher binding energies from Y to YH_2.1 to YH₃. Taking into account different core-hole screening mechanisms in these materials, however, a charge transfer from Y to H is evident in YH₂, but no clear further Y to H charge transfer is seen in going from YH₂ to YH₃. Importance of the Y 5sp-H 1s bond is suggested from the valence-band spectra.

The calculation of LEED intensities in a spherical-wave representation can be substantially simplified by symmetry relations. The wave field around each atom is expanded in symmetry-adapted functions where the local point symmetry of the atomic site applies. For overlayer systems with more than one atom per unit cell symmetry-adapted functions can be used when the division of the crystal into monoatomic subplanes is replaced by division into subplanes containing all symmetrically equivalent atomic positions.

Previous LEED results (see Maglietta et al, 1978) obtained on clean Co (001) and Co (001) c(2*2)-0 employing the spot-photometer method at low scanning rate are compared with corresponding results obtained on clean Ni (001) and Ni (001) c(2*2)-0 with a fast TV computer method. The analysis does not indicate any evidence of a significant influence of the speed of data collection on the agreement attainable with theoretical calculations, thus confirming the reliability of classical methods, and showing that they compare well with fast methods, at least for systems as stable as those examined. Cross-comparison between results obtained for very different experimental conditions also enhances confidence in the reproducibility of the experiment.

The atomic structure of the clean surface of BCC Fe(211) was analysed by low-energy electron diffraction (LEED). Thirty intensity-energy spectra (19 non-equivalent) for three directions of incidence were collected and compared with calculations by means of a reliability factor. The surface structure consists of a first-layer registry shift of 0.24+or-0.03 AA with respect to the second layer, a second-layer registry shift of 0.037+or-0.03 AA with respect to the third layer in the direction opposite to the first-layer registry shift, a first interlayer spacing d₁₂=1.05+or-0.03 AA (bulk value is 1.17 AA), a second interlayer spacing d₂₃=1.23+or-0.03 AA and a third interlayer spacing d₃₄=1.15+or-0.04 AA. The first-layer registry shift is along a (111) direction such as to decrease the difference between nearest and next-nearest neighbour bond lengths among atoms in the top two layers. The overall agreement between theory and experiment as measured by the reliability factor of Zanazzi and Jona (1977) is 0.111.