Table of contents

This volume contains the proceedings of the 14th General Conference of the Condensed Matter Division, GCCMD-14, held on 28–31 March 1994, at the Faculty of Civil Engineering of the Polytechnical University of Madrid. The publication contains the Plenary and Invited Lectures of those authors who agreed to publish their presentations. The meeting was organized by the Spanish Vacuum Society, ASEVA, under the auspices of the Condensed Matter Division of the European Physical Society, CMD-EPS.

The Conference was attended by 466 participants mostly from Europe. The emphasis of the Conference was mainly on:

• Semiconductors and Insulators

• Surfaces and Interfaces

• Liquids and Statistical Mechanics

• Magnetism and Metals

• Macromolecules and Chemical Physics

The 554 contributions were presented as 6 plenary lectures, 67 invited lectures, 140 oral presentation and 341 poster presentation, in five parallel sessions.

The guest Editors are grateful to those authors who sent their contribution for the publication, to the Organizing Committee, to the International Advisory and Programme Committee and to the Local Committee for their excellent work.

We also wish to thank those colleagues who took on the hard task of helping in refereeing the papers. It is also a pleasure to thank the Physica Scripta Editor and Editorial Board of Physica Scripta.

Electron or positron storage rings are the most powerful sources of ultra-violet and x-ray photons available to experimentalists interested in spectroscopic and structural studies in condensed matter. The most recent generation of synchrotron radiation sources is characterized by an increase by several orders of magnitude in brilliance. The potentialities opened up by these new machines are illustrated by the results of recent experiments at ESRF in Grenoble.

More sophisticated III-V devices and the challenge of integration require material growth over structured surfaces. Of particular interest are the Ga-In-As-P materials. Due to the surface selective growth in metalorganic molecular beam epitaxy a high degree of perfection in locally grown structures can be achieved. However, a clear picture of the growth mechanism on various crystal planes and structures helps optimizing the so-called selective area epitaxy. These mechanisms, in respect to relevant epitaxial parameters (V-III-ratio, rate, crystal orientation) with emphasis on lateral butt coupling structures into various crystal directions is discussed. Application of in situ grown lateral heterojunctions for e.g. laser-amplifier wave-guide integration are presented.

A theoretical analysis of single electron charging processes in a quantum dot with under a perpendicular magnetic field is presented in a broad range of the number of electrons. Exact diagonalizations and an unrestricted Hartree-Fock approximation have been used in the low and intermediate field regime to study the addition spectra and the ω → 0 spectral weight function, which are the magnitudes directly connected with single electron capacitance spectroscopy and tunneling experiments. In the high field regime the quantum dot undergoes a set of integer charge edge excitations. The results of an exact calculation suggest a simple approach for the study of these excitations.

Experimental investigations of self-generated and externally driven non-linear current oscillations due to impact ionization of shallow impurities in n-type GaAs at low temperatures are presented. The regular relaxation oscillations which appear at the onset of breakdown are destabilized by a magnetic field normal to the epitaxial layer and multifrequency oscillations and chaotic fluctuations following the Ruelle-Takens-Newhouse scenario can be found. Driving the self-generated oscillations with an external periodic bias voltage quasiperiodicity, mode locking, interaction and overlap of mode locked resonances and period doubling can be observed depending on frequency and amplitude of the external force. The reconstructed return maps, recorded phase diagrams and Devil's staircase agree with theoretical predictions based on the circle map and give new evidence of the universality of the transition from quasiperiodicity to chaos.

This paper describes how the combined use of optical spectroscopy and RBS/Channelling techniques is useful to investigate the optically active M³⁺ centres in LiNbO₃:M crystals (M = Pr, Nd, Eu, Ho, Er, Tm and Yb). Particular emphasis is made on the LiNbO₃(MgO):Nd laser system. Possible structures for M³⁺ centres are discussed in terms of different off-centred positions from regular lattice sites for the M³⁺ ions. The effect of MgO codoping (necessary for laser applications) on the M³⁺ centres is also analyzed.

Many-body Coulomb interactions dominate nonlinear optical properties of semiconductor structures in many respects. Resulting effects show up most distinct in the temporal evolution of transient four-wave mixing signals. The simplest example in systems with homogeneously broadened absorption is the static exchange interaction, which prevents the occurrence of pure free polarization decay and gives rise to an additional interaction induced signal. For inhomogeneous broadening one obtains photon echo-like behaviour. The temporal duration of the signal, however, is not determined by the width of the inhomogeneous broadening, as is predicted by theory if interaction effects are neglected, but is much smaller due to exciton-exciton interaction. Additional to the static interaction effects electron-electron correlations give rise to laser induced renormalization and dephasing processes which depend sensitively on the polarizations of the driving fields. These effects are analyzed on the basis of quantum-kinetic equations, which allow to describe the temporal evolution of the nonlinear polarizations. Numerical results are presented for various excitation conditions and material systems.

Interface roughness and layer thickness fluctuations in superlattices and quantum wells cause a partial breakdown of crystal-momentum conservation in Raman scattering for wave vectors both along (q_z) and perpendicular (q_||) to the growth direction. In the acoustic-phonon regime this leads to a continuous emission spectrum which reflects features of the foldedphonon density of states. Dispersion gaps at centre and edge of the mini-Brillouin zone as well as internal gaps at anticrossings of longitudinal and transverse phonon dispersions appear as pronounced peaks and dips superimposed on the emission background. We analyse the strength of these features in terms of an in-plane scattering potential and obtain lateral length scales of interface growth islands. From the comparison of crystalmomentum conserving folded-phonon scattering with the continuous emission signal we determine homogeneous and inhomogeneous broadenings of the electronic structure.

A theory for long-wavelength optical phonons in quantum wires and quantum dots is developed under very general conditions. Confined and interface modes are obtained as solutions of the full system of four coupled second-order differential equations for the vibrational amplitudes and the electrostatic potential. These equations are solved without incompatibility between the mechanical and electrostatic matching boundary conditions. The eigensolutions and eigenmodes for cylindrical quantum wires and spherical quantum dots are found in terms of the longitudinal and transverse parts of the mechanical vibrational field and the electrostatic field, all appropriately coupled. The pure surface phonons (Fröhlich phonons) are obtained as a particular case. The orthogonality and the general conditions under which the set of eigensolutions form a complete basis are established. The electron-phonon interaction Hamiltonians are given in explicit form.

Impurity bound magneto-polarons are studied in bulk and in weakly and strongly-coupled GaAs/AlGaAs superlattices. A variational approach is used to calculate the energy levels of several donor states as function of: (1) the well width, (2) the donor position, and (3) the magnetic field strength. The tunneling of the electron in the adjacent wells of the superlattice is incorporated in our calculation. The 1s → 2p₊ transition of shallow donors is tuned through resonance with the GaAs optical phonon mode. The theoretical analysis shows that: (1) the size of the magneto-polaron effects is very close to those for the bulk system, and (2) no direct evidence is found for non bulk LO-phonon interaction.

We report capacitance and far-infrared absorption measurements on quantum wires in which only the lowest one-dimensional subband is occupied. The wires are generated in special MISFET-type AlAs/GaAs heterojunctions with an interdigitated gate. Very pronounced steps in the capacitance as function of gate voltage reflect the onset of one-dimensional subbands. Additionally, in the one-dimensional quantum limit at very low electron density we observe a substructure which we interpret as a consequence of the renormalization of the effective confinement by electron-electron-interaction. Resonantly absorbed far-infrared radiation leads to a change of the sample capacitance and therefore allows us to investigate the collective excitations in the electron wires. We find that in our devices the fundamental one-dimensional intersubband resonance is tunable over a wide range up to 10.5 meV.

We have measured the magnetization of a mesoscopic single loop etched in a GaAs-GaAlAs heterojunction of high mobility. We have fabricated a SQUID detector of size comparable to the ring to achieve a good coupling. The all experimental set-up in fabricated on a single chip by electron beam lithography. We measured a typical current of 4 ± 2nA which is for the first time in good agreement with theory.

Dielectric response of incommensurate phases in order-disorder ferroelectrics is studied by means of pseudospin formalism. The complex dynamical susceptibility is calculated using the Bloch equations which are solved either by a perturbational approach or by a procedure introduced recently by Lovesey et al. We derive formulae for susceptibility as a function of frequency which display multirelaxational behaviour due to the coupling of the homogeneous polar soft mode to higher-order (with respect to their wave-vectors) modes.

In this paper we review the recent applications of our original tight-binding scheme to total-energy calculation of nickel-, cobalt- and iron-silicides. Both bulk and pseudomorphic (i.e. substrate-stabilized) phases have been considered, showing that the latters are actually few tenths of eV per formula unit higher than the formers. Elastic constants and optical phonons have also been evaluated in good agreement to existing experimental data.

We have developed a new technique, which allows to pattern thin evaporated Au films at the nanometer scale with a scanning tunnelling microscope (STM). The STM tip is used to expose a very thin layer of ω-tricosenoic acid, which has been deposited on top of the Au films using the Langmuir-Blodgett technique and acts as an electron sensitive, negative resist. We have fabricated narrow Au lines with a width down to 15nm and we have checked that our STM lithography does not degrade the metallic properties of the Au films. The power of our lithographic patterning technique is nicely illustrated by the fact that we can attach electrical gold contacts to a small bundle (total diameter of 50nm) of carbon nanotubes ("Buckey tubes") and measure the electrical resistance of the bundle as a function of temperature and magnetic field.

Ballistic Electron Emission Microscopy (BEEM) can be a versatile spectroscopic technique to investigate electron scattering phenomena during transport across metal-semiconductor (M-S) interfaces. Two examples for obtaining numerical values of scattering parameters are discussed. In the first example the elastic and inelastic mean free paths of electrons in Pd are deduced over a 1-6eV energy range from model fits of the attenuation of the collector current with film thickness for thin Pd films deposited on Si(111) and Si(100) substrates. The results are used to demonstrate that electron scattering in the Pd film is insufficient to account for the spectral similarities for the two Si substrate orientations, which implies that the scattering occurs at the M-S interface and that transverse momentum is not conserved for electrons crossing the M-S interface. In the other example, the impact ionization quantum yield for electron-hole pair generation in Si is directly measured over an energy range from 1-7eV by injection of electrons through pin holes in thin NiSi₂ layers grown epitaxially on Si(111). The quantum yield is in excellent agreement with existing model calculations, which can be used to determine the primary component of the collector current for any energetic BEEM spectrum. The primary component is shown to exhibit structure that is interpreted as arising from density-of-states effects in the semiconductor.

1993 has been full of landmarks for interface research: the new ultrabright soft x-ray synchrotron sources ALS-Berkeley, ELETTRA-Trieste and SRRC-Taiwan, and the user operation of the free electron laser (FEL) at Vanderbilt. They are particularly important for the increasing use of lateral resolution in photoemission, and for the first use of FEL spectroscopy. We review some recent results: the discovery of lateral band-bending changes of cleaved III-V's; [1] Gozzo et al.'s observation of spatially-varying barriers at Schottky [2] and heterojunction [3] interfaces; Cerrina et al.'s photoelectron micrographs of homojunctions; [4] Coluzza et al.'s evidence that many of the conventional photoemission spectra of insulators are heavily affected by local charging [5]. As to the FEL, we will present the two-photon absorption at the indirect Ge gap [6], and Coluzza et al.'s accurate and reliable measurements of energy barriers at "buried" interfaces with internal photoemission [7].

A generalized wave equation containing production and dissipation of energy is derived in a heuristic fashion so as to have in the dissipationless limit the (two-way wave) Boussinesq equation, while for the slowly evolving in a moving frame (one-way) wave system, it reduces to the dissipation modified KDV equation (KDV-KSV) with the same energy-balance law. The new equation allows investigating the head-on collision of dissipative localized structures. A special difference scheme is devised which faithfully represents the balance law for energy. The numerical simulations show that if the production-dissipation rate is of order of a small parameter, the coherent structures upon collisions preserve their localized character and within a time interval proportional to the inverse of the small parameter they behave like (imperfect) solitons. The collisions are almost ideal without phase shift. The only difference from the strictly soliton collision is that during the time of interaction the dissipative structures are "aging" and changing their shapes.

The traditional soliton theory deals with integrable dynamical systems possessing an infinite number of conservation laws. However, one observes solitary waves also in physical systems that are not conservative. Such "solitons" can be generated spontaneously by an instability of the homogeneous state. In some cases, this phenomenon takes place in nearly integrable systems (e.g., described by slightly perturbed Korteweg-de Vries equation), but solitary waves can exist even in essentially non-conservative systems. The present paper gives a review of some recent investigations of solitary waves and localized states in unstable systems. The main attention is granted to the instabilities in fluids produced by buoyancy and thermocapillary effects. The case of the perturbed Korteweg-de Vries equation is considered in detail.

Density stratification and wind shear provide background conditions conducive to wave propagation. Typical waves propagate by a coupling of the velocity field and the density field and are called internal gravity waves. When the shearing dominates the stratification (i.e. the global gradient Richardson Number Ri « 1) internal gravity waves are known to be strongly attenuated even at high Reynolds number. The type of waves that persist have different propagation dynamics. With small Richardson number, to a first approximation, the transport equations for thermal energy and momentum decouple. Thus, a large amplitude temperature wave has a small effect on the fluid velocity. From a dynamical systems viewpoint, dissipation attracts the velocity field toward a time asymptotic state driven by the passing temperature wave. However, transients are only weakly attracted when the forcing is strong and may oscillate if the dissipation is subcritical. The elements of nonlinearity, dissipation, and forcing are a prescription for the possibility of chaotic attractors.

Drop motions may be a consequence of surface stresses induced by the Marangoni effect (i.e. by the variation of surface tension with temperature or surfactant concentration). In the present paper we consider a steady drop motion in a homogeneous medium, when Marangoni instability associated with radial temperature and/or concentration distributions outside and inside the drop occurs. Attention is paid to the case when the instability thresholds for two different modes are close to each other and a weakly nonlinear analysis is developed.

We have calculated the combined effects of the electrostatic and hydrodynamic interactions on the short-time dynamics of charge-stabilized colloidal suspensions, which can be investigated experimentally from the first cumulant of the dynamic scattering function. The first cumulant is obtained using two methods, a pairwise-additivity approximation for the many-body hydrodynamic interaction combined with a far-field expansion of the mobility tensors, and a renormalized density fluctuation expansion due to Beenakker and Mazur. An effective hard-sphere model is employed to provide a simple physical interpretation of the results derived from both methods. It is shown that even for moderately charged colloidal particles the hydrodynamic interaction cannot be neglected, and that its effects are more pronounced for collective diffusion than for self-diffusion. Moreover, our calculations predict a non-analytic volume fraction dependence of the sedimentation velocity, with an exponent whose numerical value is sensitive to the amount of added electrolyte.

This contribution is about three related subjects:

(i) The effect of shear flow on long ranged critical correlations. An expression for the shear rate dependent structure factor is duscussed, together with the resulting anomalous behaviour of the turbidity. The predicted scaling behaviour of the turbidity, comprising both the effect of temperature and shear rate, is in good agreement with experiments on binary fluids.

(ii) The effect of shear flow on initial spinodal decomposition kinetics. An explicit expression for the time and shear rate dependent effective diffusion coefficient is discussed, which shows all the typical characteristics of anisotropic light scattering patterns that are observed experimentally for binary fluids.

(iii) The effect of shear flow on the location of the spinodal, c.q., the critical point, and the cloud point curve. It is shown that the critical point is shifted linearly with the Peclet number for not too large Peclet numbers, whereas the cloud point is singularly displaced into the unstable region on application of a shear flow. It is argued that light scattering is useless to determine the location of the critical point of a sheared system. This is in agreement with an experiment on a two polymer-solvent mixture.

We present a comprehensive study of low-temperature magnetic relaxation in magnetic systems and high-T_c superconductors (HTSC). It is argued that relaxation experiments are ideally suited for the observation of magnetic quantum tunnelling since they automatically provide the condition of very low energy barriers. Our data demonstrate the existence of two relaxation regimes: At high temperatures, the magnetic viscosity depends on temperature in accordance with the theoretical expectation for thermally activated processes. At low temperatures the viscosity is temperature-independent, providing evidence for quantum tunnelling of the magnetization. Qualitative agreement between theory and experiment is found.

The proximity effect influences the transport properties of normal metal-superconductor heterostructures. If these systems are small they exhibit pronounced charging effects. We show that charging suppresses the proximity effect. The effect can be made visible since the charging energy and hence the proximity effect can be modulated by gate voltages. We derive these results in a unified formulation which also describes single electron tunneling, Andreev reflection and the supercurrent through normal-superconductor systems in the presence of charging effects.

The irreversibility line has been measured in high magnetic fields. A new regime appears above 10T. Relaxation of the magnetization shows both logarithmic and non-logarithmic behaviours depending on B and T. A new transition line separating these two regimes is found below the irreversibility line.

We report some of the results that we have obtained from the study of the electromagnetic properties of the electron-doped L_2-xCe_xCuO₄ (L = Pr, Sm, Nd) high temperature superconductors. We will review on: intrinsic properties of the superconducting state, the flux dynamics and energy dissipation mechanisms and the critical temperature and pressure effects. We end with comments on new experiments on the thermodynamical stability of these phases.

ESR studies of the excitation spectrum of Haldane spin systems are presented. We performed temperature dependent high-field and high-frequency ESR measurements up to 21 Tesla and 1042 GHz on the compound NENP [Ni(C₂H₈N₂)₂NO₂ClO₄]. Based on extensive experimental investigations we propose to distinguish between different classes of resonances: first, the transitions between excited states at the boundary of the Brillouin zone (q = π) below the critical field; second, the high-frequency transitions from the ground state (q = 0) above the critical field and third, the low-frequency resonances at q = 0 in the whole field range. New results are presented in the high-field phase, which exhibit an appreciable temperature dependence of the resonance field and at low temperatures show the existence of multiple high-frequency excitation branches. In the low-frequency regime we also observe an additional ground state excitation.

The magnetization of three different systems of diluted easy-axis antiferromagnets, Mn_1-xZn_xF₂, K₂Fe_1-xIn_xCl₅.H₂O and Rb₂Fe_1-xIn_xCl₅.H₂O, reveals that a remanent magnetization, M_r, developes below the Néel temperature. The remanent magnetization is observed when fields as low as 10^-3 Oe are applied parallel to the easy axis. At a given temperature, M_r increases with H_axial but is practically saturated in fields as low as a few Oe. The normalized remanent magnetization M_r(t)/M_r(0) is an universal function of the reduced temperature t = T/T_N. The function is the same for all three site-diluted systems and is independent of x and H_axial. The origin of M_r is still not known with certainty, but several experimental results suggest the formation of a domain structure, possibly with magnetization in the domain walls. The possibility of the piezomagnetic effect as the origin of M_r, as suggested by other authors, is excluded on the basis of symmetry arguments and the observed magnetic structures of K₂FeCl₅.D₂O and Rb₂FeCl₅.D₂O.

Spontaneous nuclear magnetic ordering is discussed for the following two groups of metals:

1. Metals where the nuclear-electron interaction is so weak that the nuclear spin temperature can be reduced substantially below the electronic and lattice temperatures and where this temperature difference can be kept for a time sufficiently long to perform investigations of the "isolated" nuclear spin system. By this method, nuclear magnetic ordering in Ag and Cu has been observed at nanokelvin nuclear spin temperatures. Here I will also consider the interesting phenomena occurring at negative nuclear spin temperatures.

2. Metals where the nuclear-electron coupling is so strong that nuclei and electrons cannot be decoupled for an experimentally relevant time. For these groups of metals the complex NMR spectra of Tl as well as the recent observation of a spontaneous nuclear magnetic ordering transition of In nuclear spins in the cubic intermetallic compound AuIn₂ at microkelvin temperatures will be reviewed.

The superconducting field (H)-temperature (T) phase boundary has been measured in mesoscopic Al samples of different topology: lines, open and filled squares, which were made under the same conditions from the same material. These samples clearly show different superconducting H-T phase boundaries which are nicely reproducing the predictions of the theoretical calculations made for their particular confinement geometries. The confinement of the flux lines by the lattice of the submicrometer holes has been studied in the Pb/Ge multilayers. A substantial enhancement of the critical current j_c has been achieved. Sharp integer and rational matching peaks in the j_c(H) curve are observed. The possibility of the "quantum design" of the superconducting critical parameters (H_c(T) and j_c(T, H)) of the mesoscopic and nanostructured superconductors by optimizing the confinement geometry for the superconducting condensate and for the flux lines has been demonstrated.

Density Functional Theory (DFT) is used to study the structure of bimetallic clusters. Aggregates formed by two or three different alkali elements are first considered and the Kohn-Sham version of DFT is employed to analyse mixing and segregation properties, and their influence on the collective electronic response of the cluster. Then Cu-Ni clusters are chosen as an example of non simple metals. In this case the embedded-atom method, rooted on DFT, is used to study the structure of these clusters.

A polaron-type model is proposed to describe the deformation in the superfluid ⁴He ground state, due to the presence of a ³He atom. The interaction matrix element of the model is expressed in terms of the structure factor of the superfluid on the basis of the hypernetted chain approximation.

Both the use of an intermediate polaron coupling theory (rather than perturbation theory) and the determination of the precise interaction matrix elements are crucial ingredients to obtain an adequate theoretical description based on a polaron-type model. A generalization of Feynman's path-integral technique is used for the treatment of intermediate coupling.

The present work solves i.a. an outstanding problem regarding the ³He effective mass in superfluid ⁴He. The theoretical result (an effective mass of 2.3123m₃ as compared to the experimental value of 2.3m₃ where m₃ is the bare ³He mass) confirms that our formalism accurately describes the deformation cloud around the ³He atom.

Also the difference between the specific heat of the mixture and the specific heat of pure superfluid ⁴He, and the diffusion of ³He in the superfluid, are accurately explained as the result of the interaction of ³He with the excitations of the superfluid.

We have fabricated pillar-like microstructures of Fe/Cr and Co/Cu magnetic multilayers and measured the giant magnetoresistance effect with the current perpendicular to the multilayer plane. The perpendicular giant magnetoresistance effect was determined as a function of temperature from 4 K to 300 K. For both Fe/Cr and Co/Cu multilayers we find, at low temperature, magnetoresistance effects of the order of 100%. The Fe/Cr pillars show a pronounced decrease with temperature of the magnetoresistance, while for Co/Cu the temperature dependence is much weaker. This different behaviour is caused by the different temperature dependence of the spin-dependent scattering; in the case of Co/Cu, we present results of a quantitative analysis of the temperature dependence using a resistance model of Fert and Campbell, originally proposed for magnetic alloys [1].

Fe-rich amorphous alloys containing late transition metals like Nb, V, Zr,..., sometimes with the addition of Cu, can crystallize in ultrafine grains of a crystalline phase, a few nanometers in diameter, embedded in a disordered matrix. In such state they have shown excellent soft magnetic properties for technical applications, rising the interest for deep studies. In this paper, recent work on some Fe-Nb and Fe-Zr based alloys both in amorphous state and after several degrees of nanocrystallization is presented. The nanocrystallization process has been achieved by conventional heat treatments (about 1 h at temperatures around 400-500 °C in a controlled atmosphere furnance) as well as by Joule heating using an electrical current flowing through the sample. Magnetic measurements, electrical resistivity, x-rays diffraction and ⁵⁷Fe Mössbauer spectroscopy were used in the study of the crystalline phases appearing after the thermal treatments. The basic magnetic and transport properties of the nanocrystals do not differ appreciably from their bulk values. The magnetic anisotropy, however, is very sensitive to grain size and to the intergranular magnetic coupling. The effect of such coupling is deduced from the coercivity changes at the Curie Temperature of the amorphous matrix remaining after nanocrystallization.

Techniques now exist that allow near-surface mechanical and thermal properties of polymers to be measured at very high depth resolution. We review the use of: nanoindentation (hardness, modulus, flow properties, profilometry); force spectroscopy (adhesion, attractive forces and their variation with separation); and scanning calorimetric microscopy (image contrast determined by spatial variation of thermal conductivity, heat capacity, or reversing/non-reversing behaviour, as a function of temperature).

Polymeric materials present a major challenge to computer simulation, since they exhibit structure on many different length scales (bond length, 1 Å; coil size, 10² Å; correlation length in mixtures, 10³ Å). For a study of collective phenomena, such as unmixing of polymer mixtures and mesophase formation in block copolymers, coarse-grained simplified lattice models hence are used. It is shown that in symmetric polymer mixtures (chain lengths N_A = N_B = N) the critical temperature follows a simple linear relation, T_C(N) = aN + b. The onset of crossover phenomena from Ising to mean field critical behavior is observed, but even for N = 512 chains do not behave fully mean-field-like. Asymmetric mixtures also are considered and the corresponding predictions of Flory-Huggins theory are tested. In block copolymers a pretransitional stretching of the chains is found for both symmetric and asymmetric concentrations and analysed in detail. A density reduction in the interfaces between A and B blocks is found in the ordered phase.

The change of the α relaxation process occurring during isothermal crystallization of a glassy polymer, poly(aryl ether ketone ketone), has been followed in real time through measurement of the dielectric complex permittivity. The phenomenological Havriliak-Negami description has been used to analyze the changes of the dielectric strength, central relaxation time and shape parameters describing the relaxation, as crystallization proceeds. The evolution of the dielectric magnitudes with crystallization time has been interpreted in the light of the Schönhals-Schlosser model. According to this model, a restriction of the long scale motions of the polymeric chains as the material is filled in with crystals is observed. These results may contribute to the better understanding of dynamics of a glassy polymers during crystallization.

In this paper, some of the essential relationships between the molecular structure of polymer networks and their macroscopic mechanical properties are outlined and referenced. Special attention is given to the deformation behaviour of physical networks and to the competition between crazing and homogeneous deformation in glassy thermoplastics.

The use of thermodynamics of irreversible processes allows one to describe time dependent deformation phenomena of polymer networks. It is shown that linear relaxation modes are coupled to the network ("global level"). The constitutive van der Waals equation is characterized by a strain, strain-type and a strain-rate independent memory function. Stress-strain cycles in the glass-transition regime can be described.

The dynamics of networks can be characterized by introducing the concept of conformons. Conformons are shown to be Bosons with a photon statistics. Conformons in equilibrium have a self-energy- and self-entropy-term. They can be excited or emitted. Conformons also decay spontaneously. In the stationary limit a broad conformon spectrum results which is typical for glass forming liquids (including polymers). Relaxation-and retardation-time spectra are shown to be related to the conformon spectrum. Mechanical and dielectric relaxation measurements (even of low molecular weight glass formers) are fairly well reproduced. The dynamical structure factors of the incoherent neutron scattering can also be computed.

In early research on polyethylene terephthalate, studies of shrinkage force, shrinkage and birefringence on melt spun fibres confirmed the existence of a molecular network which was of key importance in determining the strain hardening behaviour and hence the natural draw ratio. These ideas were translated to the ultra drawing of polyethylene, where it was found that molecular weight and initial morphology determined the ultimate extensibility of the polymer. The existence of the network in the drawn polymer was confirmed by measurements of thermal expansion and shrinkage force.

Recently this research has been extended to define the network which forms in the spinning threadline for very high wind up speeds and how this subsequently affects the drawing behaviour and the final properties of the drawn yarns. Specific issues addressed have included the effect of molecular weight on the behaviour of polyethylene terephthalate fibres and the extension of the studies to polybutylene terephthalate where changes in the crystalline structure can occur.

In films, a comparison has been made between uniaxial and uniaxial planar (constant width) films. Refractive index and infrared spectroscopic measurements have been undertaken to determine the chain axis orientation and conformational changes. It has been shown that a quantitative description can be obtained in terms of the deformation of a rubber-like network. More recently, studies have been undertaken of two-stage stretching, with the additional objective of determining the factors influencing the development of crystalline orientation in the second draw direction.

Erratum to Physica Scripta, T49B, 706 (1993).