Table of contents

First a brief description is given of the experimental techniques used for ultrasonic investigation of solids, in particular the use of wavevector reversed acoustic signals, a method developed recently in our group. Examples are then discussed of results obtained, experimentally and theoretically on the behaviour of sound attenuation and velocity near structural phase transitions. Particular emphasis will be placed on the discussion of results in perovskites (ABX₃), the most thoroughly studied group of materials. Recent work in our group has shown that critical attenuation in KMnF₃ obeys dynamic scaling, a feature not yet found in other materials. Since the ultrasonic frequencies employed are only a few hundred megahertz a very slow relaxation time τ must be invoked here. This time constant is interpreted as the characteristic relaxation time of ordered clusters of range ξ. Analysis of the results led to critical exponents in agreement with renormalization group calculations, also implying agreement with the predicted value of the crossover exponent: ϕ = 1.25. Recent work in our laboratory on the direct measurement of ϕ will also be discussed.

The spin density waves of antiferromagnetic chromium are characterized by a transversal polarization of the spins between the Néel point and the spin-flip temperature T_sf while for temperatures below T_sf the spins are longitudinally polarized. The transition at T_sf is of first order. We report here direct measurements of the latent heat of this transition on a vapour-grown chromium crystal using a differential calorimetric technique. The latent heat determined is 0.04 ± 0.02 J/g-atom and T_sf is 123.2 ± 0.4 K. A thermal hysteresis of 0.8 deg. was observed at the transition. The magnitude of the latent heat correlates well with the pressure dependence of T_sf through the Clausis-Clapeyron relationship.

Even for pure, well annealed specimen experimental data show rounding of the cusp in c_p and dR/dT at the Curie temperature. This effect has generally been attributed to structural defects (dislocations, grain boundaries), but no quantitative explanation has been given so far. In connection with the analysis of very high precision resistance measurements on nickel it was discovered that structural effects could only account for a minor part of rounding, the main part being caused by critical slowing down, i.e., the increasingly slow relaxation towards new equilibrium values for the resistance, when T_c is approached. Reanalysis of published data on the specific heat of nickel has revealed slowing-down effects of the same nature.

Based on the divergent behaviour of the relaxation time τ at T_c, τ ∝ T/T_c - 1^-Δ, a semi-quantitative explanation is given for the rounding and an approximative value for Δ obtained.

In a SQUID magnetometer we have simultaneously measured the a.c. susceptibility (6 Hz, 10 nT (0.1 mG) rms) and the d.c. magnetization at applied d.c. fields up to 0.5 mT (5 G) of the amorphous materials (Fe_0.20Ni_0.80)₇₅P₁₆B₆Al₃ and (Fe_0.68Mn_0.32)₇₅P₁₆B₆Al₃.

Both samples have an iron content slightly higher than the concentration at the multicritical point for the respective alloy system. When the temperature is increased, the samples are characterized by spin-glass (SG), ferromagnetic (FM) and paramagnetic (PM) behaviour.

In both samples the FM-PM transition is well defined (Fe, Ni): 87 K; (Fe, Mn): 98 K, while the SG-FM transition is ambiguous. In the (Fe, Ni) sample a rather broad SG-FM transition appears at approx. 20 K in both a.c. and d.c. measurements. The d.c. thermoremanent curve (TRM) shows a minimum at the same temperature. The (Fe, Mn) sample shows a sharp decrease for decreasing temperature at 57 K in a.c. as well as d.c. measurements, suggesting the position of the SG-FM transition. The TRM curve shows a pronounced dip at the same temperature and typical viscosity effects are observed below this temperature.

Fe₂P (space group P2m) is a highly anisotropic, metallic ferromagnet with the spontaneous magnetization directed parallel to the c-axis. The ferromagnetic transition at T_c = 217 K is of first order and accompanied by a sudden change in lattice parameters.

In order to study the ferromagnetic transition we have measured the specific heat of an annealed powder sample in a ³He-cryostat in the temperature range 2-20 K and in a liquid helium cryostat in the range 80-300 K, in both cases using helium as heat exchange gas.

The low temperature measurements give an electronic part of the specific heat characterized by γ = 24 mJ/K² · mole and a lattice part with a Debye temperature of Θ = 420 K at the lowest temperatures. Effects of helium adsorption on the powder sample (surface area ≈ 0.5 m²) were accounted for.

The ferromagnetic transition is only noticeable through a sharp, symmetric peak at 217 K (width 0.6 K) raising above a smooth background. The corresponding entropy change is ΔS ≈ 0.2 J/K · mole (≈ 0.02 R) in reasonable agreement with an entropy change calculated from Clapeyrons equation.

The effect of uniaxial stress on the magnetization of a ferromagnetic single crystals of cubic structure is calculated by minimizing the free energy. Calculations are made for the case of a specimen in the shape of an oblate spheroid parallel to a (100)-plane, with field and stress applied in arbitrary directions in the plane of the specimen. The magnetization curves obtained show three separate regions. Analytical expressions are derived for the first part of the magnetization curve, the two other parts are calculated by means of numerical methods.

Squaric acid is a layered crystalline material with squaric molecules bonded by O-H... O-bonds, and showing a structural transition at 100°C. Recently Raman studies and incoherent neutron work have confirmed the earlier indication of local order being retained into the high-temperature phase, but new evidence from coherent neutron work points to a one-dimensional-like rather than a two-dimensional correlation above T_c. The analysis of the data may point to a soliton-like type of excitations.

The Zeeman splitting of the conduction electrons has been investigated by means of the de Haas-van Alphen effect. Utilizing a magnetic field of 14 T made it possible to study the amplitude variation of the second harmonic of the de Haas-van Alphen signal for large portions of the Fermi surface. Several spin-splitting zeros of the second harmonic were located.

The Zeeman splitting of the α-orbit is found to follow the effective cyclotron mass anisotropy and thus the g-factor is isotropic. On the contrary substantial anisotropy is found on the closed Γ6 sheet for cyclotron orbits in the vicinity of the [100]-axis. The Zeeman splitting is observed to change as much as one half Landau level spacing when the field direction is changed 2°.

The cohesive properties of heavy-element materials have been calculated by means of self-consistent relativistic LMTO procedures in conjunction with the local-density approximation to exchange and correlation. The importance of including the relativistic effects-also the spin-orbit coupling-in the self-consistency scheme is illustrated by taking Xe and Au as examples.

Dielectric losses and dispersion have been studied from 30 kHz to 15 000 MHz at low temperatures in many ammonium salts. In (NH₄)₂SnCl₆ and NH₄Al(SO₄)₂ · 12H₂O loss maxima frequencies coincide with those deduced from proton NMR and NH₄⁺ reorientation. This shows that NH₄⁺ in these salts reorients by hopping between two positions with different centres of charge. The NH₄⁺ motion may be important for ferroelectric phase transitions.

Brillouin spectroscopy has been used to investigate the acoustic excitations in [N(CH₃)₄]₂CuCl₄. This crystal is known to have successive transitions from the prototype phase (D_2h¹⁶) to the incommensurate phase at T₁ = 297 K, to the ferroelastic commensurate phase (C_2h⁵) at T₂ = 291 K and to another monoclinic phase at T₃ = 263 K. Upon transition to the incommensurate phase no anomaly could be detected in the velocities of the longitudinal phonons propagating in the [100], [010] or [001] directions. Within the incommensurate phase the velocity of the mode propagating along [001] showed a decrease of about 1.2%. The Brillouin line width of this mode was found to depend on temperature both in the incommensurate and in the commensurate phases, attaining the maximum value near T₂. The results are compared with the existing ultrasonic data and discussed by assuming the phason-phonon interaction in the incommensurate phase.

An experimental system is discussed where hydrogen atom fluxes of up to 10¹⁶ H s^-1 are fed into a liquid He cryostat and are detected using recombination calorimetry.

Recent progress in electronic structure and related properties of point defects in metals is reviewed. Topics discussed include transition metal impurities in a non-magnetic host, construction of potential energy surfaces from effective-medium theories, trapping of light interstitials, and pair potential in metals.

ΔSCF density-functional calculations are reported for Auger, and core level binding energy shifts in sp-bonded metals. The basic model, atom-in-jellium-vacancy, gives good agreement with experiment, especially in the Auger case. The chemical and relaxation contributions to the shifts are discussed. The shifts are calculated also by using the thermochemical model and the results obtained are in agreement with experimental data. The applicability of the "excited-atom" approach to the Auger energy shifts is found restricted.

The magnetic properties of microcrystals are discussed on the basis of results obtained by use of Mössbauer spectroscopy. Spectra of microcrystals with dimensions smaller than about 100 A are influenced by magnetic fluctuation effects such as superparamagnetic relaxation and collective magnetic excitations, and these effects allow a determination of the product of the particle size and the magnetic anisotropy energy constant. Measurements on ferromagnetic and ferrimagnetic particles as a function of applied magnetic fields allow a determination of the particle volume. The anisotropy energy constants of microcrystals of Fe, Co, Ni and Fe₃O₄ are found to change upon chemisorption. Spectra of small particles at low temperatures also give information on the magnetic properties of the atoms in the surface layer. These studies as well as thin film studies show that no "magnetically dead layers" are present at the surface of the samples. In the case of metallic iron an enhanced magnetic hyperfine field is found, in agreement with theoretical results.

The intermetallic compound Ni₃Al orders in the simple cubic (Cu₃Au) crystal structure and becomes ferromagnetic below about 40 K with a spontaneous magnetization extrapolated to absolute zero and zero external magnetic field corresponding to approximately 0.075 μ_B/Ni atom. Measurements of the first magnetocrystalline anisotropy constant K₁(T, H) have been carried out in stoichiometric single crystals of Ni₃Al in magnetic fields 0 ⩽ H ⩽ 16 kG and in the temperature range 4.2 ≲ T ≲= 42 K by means of a high sensitivity torque magnetometer. At 4.2 K and 0 kG (extrapolated from 6 ⩽ H ⩽ 16 kG) we find K₁ = - (9.6 ± 0.7) × 10³ erg cm^-3, a value having the same sign but about a hundred times smaller magnitude than that for pure Ni at the same temperature and field. The above value of K₁ has been reproduced within experimental error in three stoichiometric specimens having Curie temperatures (T_c) between 39 and 42 K and residual resistivity ratios [ρ(293 K)/ρ(4.2 K)] in the range 17 to 34.

As in pure ferromagnetic Ni, the decrease of the magnitude of K₁ with increasing temperature is, at least for T ≲ T_c/3, markedly more rapid than would be expected in terms of conventional Zener theory. The results are discussed qualitatively in relation to current models of magnetic anisotropy in itinerant electron ferromagnets and compared with recent Fermi surface studies carried out in this laboratory.

The magnetization of the amorphous (Fe_xNi_1-x)₇₅P₁₆B₆Al₃ system has been measured for iron concentrations larger than the percolation limit x = 0.17. Up to x ≈ 0.30 we find a low temperature regime of spin glass behaviour, followed at intermediate temperatures by a ferromagnetic state. For higher iron concentrations the strong exchange interactions result in ordinary ferromagnetism. For the low temperature magnetization a spin wave analysis is presented.

The influence of structural relaxation upon the ferromagnetic transition is reported as well as possible attendant consequences of the scaling law analysis. In particular the effects of annealing temperature have been studied. Despite a definite shift in transition temperature with annealing above 380 K the critical exponents are shown to be unaffected. For the pure ion sample Fe₇₅P₁₆B₆Al₃ the critical exponents (β = 0.38, δ = 4.8, γ = 1.4) are close to a 3d Heisenberg behaviour, while the mixed (Fe, Ni) samples (x < 1) consistently give exponents (β ≈ 0.40, δ ≈ 5.1, γ ≈ 1.6) that deviate from the 3d Heisenberg values.

Studies of the magnetic properties of crystals of the mixed semiconductors Cd_1-xMn_xTe indicate that: (i) for x ⩽ 0.17 the crystals are paramagnetic at all temperatures, (ii) for 0.17 < x ⩽ 0.60 a spin glass phase and (iii) for 0.60 < x ⩽ 0.71 an antiferromagnetic ordering are observed at low temperatures. Neutron diffraction experiments have shown that for 0.60 < x < 0.70 the particular peaks observed could be interpreted in terms of type III antiferromagnetic ordering of the f.c.c. sublattice. The temperature dependence of the integrated intensities of the magnetic reflections follows the square of the Brillouin function for S = 5/2, except above 3/4 of the ordering temperature. Above this temperature, the observed intensities are larger than expected from ordinary critical fluctuations near a second order phase transition. For x = 0.70, a detailed study has shown that this phenomenon suggests that the spins of the randomly distributed Mn atoms are ordered in clusters. A rather large discrepancy between the ordering temperatures obtained by neutron scattering and by other techniques is explained by a phenomenological description of the transition process.

The effect of the spin-orbit, Darwin and mass-velocity operators on the non-relativistic self-consistent valence and conduction band edges of trigonal Te is studied with the degenerate first order perturbation theory. It is found that: (1) the k dependence of all the three relativistic corrections is essential for the proper description of the band edges on the meV scale. (2) The Darwin correction cancels out a part of the mass-velocity correction to the shape of the band edges. (3) The spin-orbit and the joint Darwin-mass-velocity correction to the shape of the band edges are of the same order of magnitude. Therefore the neglect of the latter correction is not entirely justified. The calculated pressure dependence of the valence band edge of Te agrees closely with the conduction measurement of Koma et al. The experimental difference between the valence band minimum and the saddle point, Δ, by Glushkov et al. decreases about twice as fast as the calculated Δ. Therefore the α-β fine structure in the infrared absorption spectrum does not seem to be directly related to the valence band edge.

The optical reflectance and transmission spectra of GaAs at low temperatures has been interpreted in order to derive the basic parameters of exciton polaritons. Taking into account a well defined repulsive surface potential acting on the excitons and using a frequency independent life time broadening Γ it is possible to explain the observed spectra in great detail. The results are ℏΓ = 0.010 meV and the longitudinal transverse splitting ℏω_LT = 0.17 ± 0.02 meV.

In this paper, the general features of the magnetic phase diagrams, determined experimentally by a.c.-susceptibility and other techniques, for glassy metallic alloys of the type (A_xB_1-x)₇₅P₁₆B₆Al₃ (where A, B are transition elements Fe, Co, Ni, Mn or Cr) with mixed exchange interactions are reviewed. Evidence for a second-order nature of the transitions across the para-, ferro-, and spin-glass "phase"-boundaries using scaling techniques are discussed. Manifestations of the transitions across the para to ferro followed by the ferro- to spin-glass phase boundaries at lower temperatures as observed in various experimental techniques are also briefly presented.

Extended X-ray Absorption Fine Structure of amorphous and crystalline Mo-Ge samples sputter deposited on glass or kapton substrates was studied. Small local distortions were found in a substitutional b.c.c. Mo rich sample. A coordination in the range 5-7 and Ge-Mo distance of 2.65 A were estimated for an amorphous, intermediate composition Mo-Ge sample. The lack of superconductivity of some samples deposited on kapton was correlated to the presence of oxygen in the material.

The Cr-Si system exhibits above about 1 at% Si a large first order antiferromagnetic transition which is not well understood so far. Simultaneous measurements of thermal conductivity λ and electrical resistivity ρ on two Cr-Si alloys of 0.9 and 1.4 at% Si in the temperature range 3-300 K have been carried out to gain further insight into the role of electronic and lattice properties at the magnetic transition at the Néel point T_N. At T_N a step-like change in λ was observed for the Cr-1.4 at% Si alloy while λ was almost featureless for the lower concentration. The Lorenz number L, which was found to lie in the range 4-7 WΩ deg^-2, exhibits a broad peak centered around 60 K. Analysis of the λ-data reveals below 50 K a dominating lattice thermal conductivity term with a T² dependence and decreasing in magnitude with increasing Si concentration. Around T_N electronic contributions are found to dominate the thermal conductivity although both lattice and electronic effects seem to be responsible for the enhancement of L over the Sommerfeld value L₀.

For almost all amorphous magnetic systems studied so far, the critical indices for the phase transition at the Curie point are non-Heisenberg like (β ≃ 0.4, γ ≃ 1.6). Via scaling relations this gives the specific heat index α - 0.4, indicating that the singularity at T_c is considerably reduced. To confirm this α value, a direct measurement of α would be valuable. This is possible via resistance measurements, since dρ/dT has the same critical behaviour as C_P.

We have carried out precision resistance measurements on two Fe-Ni-based amorphous alloys (Fe₁₅Ni₆₀P₁₆B₆Al₃ and Fe₁₁Ni₆₉P₁₄B₆) around T_c. The shape of the dR/dT vs. T curves shows the typical cusp above T_c, but straight line below T_c, indicating suppression of the critical behaviour, and the presence of a T² term. The analysis of data give α = -0.37 ± 0.05, in fair agreement with the value calculated via scaling relations. The suppression of the critical behaviour below T_c is reflected in an anomalous value for the amplitude ratio.

The thermal expansions of some SiFe_xCo_1-x (x = 0.0, 0.5, 0.65, 0.8, 0.9 and 1.0) alloys have been measured from 4.2 to 15 K or well above their Curie temperatures using a three-terminal capacitance dilatometer. The ferromagnetic alloys in this series are shown to exhibit in their thermal expansions a negative magnetic contribution, comparable to that found in the weak itinerant ferromagnets ZrZn₂, Ni₃Al and NiPt. The present results for SiFe_xCo_1-x alloys are discussed together with the previously reported similar data on the TiFe_xCo_1-x system.

The existence of sharp extrema in the electromagnetic-acoustic coupling at the surface and in the bulk of ferromagnetic metals at specific orientations of an external d.c. magnetic field are reported here for the first time. The orientational anomalies are compared with measurements of non-anomalous acoustic resonance modes in the same samples as well as with existing theoretical formulations of electromagnetic-acoustic coupling in ferromagnetic metals.

The energy dependence of the reflection of low-energy incident electrons (range 0-10 eV above the vacuum level) from the surfaces of 1T-ZrS₂ crystals is discussed in terms of the onset of inelastic electron-electron scattering accompanied by interband transitions. Using the published density of states for ZrS₂ and the model previously applied to another dichalcogenide, MoS₂, it is possible to assign those interband transitions which give rise to prominant features of the energy structure in the integrated yield.

The performance of cooled photoconductive detectors made from high resistivity indium antimonide depends strongly on the state of the semiconductor surface. After seal-off some of the detectors show a degradation in performance due to the bake-out at 110°C in darkness during evacuation of the detector dewar. Irradiating these detectors with high-energy photons improves their performance, while subsequent heating in darkness leads to a new degradation.

A number of such detectors have been investigated leading to a theory for this radiation-dependent behaviour. Some field-effect measurements are performed to confirm the theory. The investigations on the low performance detectors also result in procedures for surface treatments giving a better yield of high performance detectors.

Selectively transmitting tin-oxide films can be produced by spraying SnCl₄-solution on a hot surface. A sufficiently high substrate-temperature T_s is required to produce optical selectivity and electrical conductivity. Detailed optical and diffraction-measurements are reported for SnO₂-films produced at different T_s.

Diffractometer-measurements on films obtained for 180 ⩽ T_s ⩽ 300°C resulted in one broad maximum, characteristic for amorphous substances. No trace of SnCl₄ was detected. Annealing these films for 1 h at 400°C resulted in a diffraction pattern typical for the structure of SnO₂. Raising T_s to 340°C resulted in the appearance of crystalline structure with a non-normal texture.

Both crystalline and amorphous films have high transmission for 0.4 < λ < 2 μm. The onset of high infrared reflectance is shown to coincide with the formation of crystallites in the film. Using diffractionpeak area as a measure of crystallinity, we give a connection between crystallinity and infrared reflectance.

The reflectivities of TiN, ZrN and HfN have been measured in the spectral energy range 0.1-6.2 eV. The optical constants have been determined by means of a Kramers-Kronig analysis. A separation into interband transitions and free electron behaviour was carried out. The extended free electron region is explained and the interband transitions are compared with allowed transitions in the density of states distribution. The possibility of improving the optical selectivity by shifting E_F is discussed.

The structural properties of the ionic conductor LiI·D₂O have been studied by neutron scattering. The cubic room temperature α-phase, Pm3m, is disordered both with respect to the occupation of the Li⁺-positions and to the orientations of the water molecules. A first order phase transition from the α-phase to a new orthorhombic β-phase, P2₁am or Pmam, has been found at - 54°C. The structural parameters of the β-phase have been determined by the Rietveld method based on powder diffraction data. The β-phase is found to have an antiferroelectric ordering of the water molecules and a spatial ordering of the Li⁺-sites. Evidence for short range order is found in the diffuse elastic scattering in the α-phase and the implications for the ionic conductivity are discussed.

The K emission band of ferromagnetic nickel has been calculated using a one-electron model and dipole approximation. The nickel K band emission was measured using a two crystal X-ray spectrometer. The FWHM of the calculated K emission band shows only about 0.1 eV broadening as compared to the calculated band of paramagnetic nickel. The experimental band is very sensitive to the correction procedure applied to the measured data but any reasonable realistic correction treatment still yields a considerably broader band than the theoretical one. This is contrary to the case of the L emission and XPS valence band for which the calculated bands are much broader than the measured ones. Consequently, the many-body effects already found in XPS and UPS spectra of nickel may also play a significant role in the Ni K band emission.

A comparison between the LEED-type (one-step) calculation and the three-step model calculation for the spectral distribution of photoelectrons at normal emission from the copper [1 1 1]-surface has been carried out. In the three-step model calculation the necessity of the proper treatment of the interpolation parameters to calculate the momenum matrix elements has been pointed out.

The basic principles of a simple apparatus for efficient recording of Bremsstrahlung in the UV-range is outlined. The radiation is analyzed with the use of a bandpass filter which is composed of a VUV window and a photocathode. The electrons from the photocathode are counted with a channeltron. Spectra of polycrystalline W obtained with LiF and CaF₂ windows are reported.

The bandstructure, densities of states and K X-ray absorption of Li have been calculated by the APW method. Comparison with a measurement of the absorption coefficient shows good agreement with respect to the energy positions of the spectral features. However, the ground state calculation fails to reproduce the relative magnitudes of the features found in the experimental absorption curve.

The K absorption spectrum of magnesium has been measured with a two crystal spectrometer and compared with bandstructure calculations. A weak structure was observed in the neighborhood of the absorption edge.

The slightly asymmetric Cu Kα₁ line has been fitted with Doniach-Sunjic lineshape function. The agreement was good down to 1.6 eV below the Fermi level. It is therefore concluded that the asymmetry is caused by core hole effects rather than by p-d exchange interaction.

The salient features of the imaginary part of the frequency-dependent permittivity ₂ of trigonal Te are interpreted in terms of the transitions between the band triplets. The interpretation is based on the self-consistent Hartree-Fock-Slater calculation where the optical matrix elements are treated rigorously. The agreement with the experimental ₂ is close. The role of the transitions from the s-bonding triplet in the higher energy region of ₂ is clarified. The uppermost p-lone-pair triplet plays the dominant role in the formation of the ₂ spectra, especially for the perpendicular component. The contribution from the p-bonding triplet falls into the energy region 4-10 eV and it is mainly responsible for the anisotropy of ₂ here.

Superconductivity is reported for a series of melt-spun Zr_xCo_1-x alloys with 0.6 < x < 0.8. The superconducting T_c is found to vary rapidly with x.

The dispersion relations for the acoustic and optic phonons in HgTe and for the acoustic phonons in HgSe were determined by neutron inelastic scattering in three high symmetry directions. The effect of the free-carrier screening of the long-range electric field of LO phonons in HgTe was observed. The formalism of the rigid ion model is used for numerical calculations of the phonon dispersion relations and the phonon densities of states in HgTe and HgSe.

Extensive measurements of radiation emitted from long Josephson junctions d.c.-current biased on zero-field and Fiske-steps are reported. The measurements were made in the band 8-10 GHz on long Nb-Nb oxide-Pb overlap tunnel junctions with the lengths 0.5 mm and 1.0 mm having fundamental frequencies of 10 GHz and 5 GHz, respectively. Depending on the fundamental frequency of the junction 1st or 2nd harmonic detection was employed. In the fundamental scheme the frequency f_n vs. the d.c.-voltage V_n along the steps follows a modified Josephson voltage/frequency relation f_n/V_n = 2e/nh, where n is the step-number n = 1, 2, 3,... for Fiske-steps and n = 2, 4, 6,... for zero-field steps. On a given step the frequency could be tuned (200-500 MHz) by varying the bias point. The linewidth was measured to be less than 5 kHz. All the observed phenomena are consistent with a novel unified dynamic bunched fluxon model applicable to both zero-field and Fiske-steps and are in good agreement with recent numerical solutions of the perturbed sine-Gordon equation with proper boundary conditions. Taking advantage of operation on higher order steps and strong harmonic generation the present upper frequency limit for the Josephson fluxon oscillator is above 200 GHz. The maximum power observed so far is only 5 pW, but the estimated available power is of the order 1 μW, which is quite sufficient to pump other Josephson devices.

Photon energies of 11.7, 16.85 and 21.2 eV were used to observe weak but strongly dispersive features in the angular photoemission in the (-10) and (-11) planes from an Al (111) surface. Structures seen in both detection planes can be excellently accounted for by direct bulk transitions of assisting reciprocal lattice vectors G₀₀₂ and G₂₂₀. The 11.7 eV spectrum for angles between 35° and 60° towards the [001] axis is highly reminiscent of that previously reported for this surface at 10.2 eV, which was interpreted in terms of a surface resonance band associated with the sp gap at the X-point. We confirm such a surface component in the spectra by discerning an emission in the opposite k_||-direction where bulk transitions are absent. Following a recent suggestion by Rosei et al., we have used Fermi surface data and our own measurements to determine the momentum k of transitions originating at the Fermi level. Agreement with standard band-structures is good at the Fermi energy, while considerable discrepancy and uncertainty exists for the final energies.

During the last couple of years catalytic reactions on Pd surfaces have been studied at our department by means of a hydrogen sensitive Pd-MOS structure and work function measurements. These studies have been performed at atmospheric pressures with Ar and O₂ as carrier gases. We have now extended this type of measurements to the ultra-high vacuum (UHV) region and also combined them with both UV and, in certain cases, X-ray photoelectron spectroscopy (UPS, XPS) studies. With this technique we are able to detect about 0.001 of a monolayer of hydrogen, and changes in rate or equilibrium constants corresponding to changes in an energy coordinate of less than 10 meV. Furthermore, results from a Cd contaminated Pd-MOS structure indicates that the ability of Pd to dissociate H₂ is related to a microscopic parameter and not to any general metal-like parameter.

The conditions for preparing selective surfaces by combinations of absorber-reflector tandems are discussed. The optical constants for Fe₂O₃, Cr₂O₃, CuO and Cu₂O have been measured and the possibility of preparing selective surfaces based on these, and some related oxides, on steel- and copper-substrates has been investigated.

Magnetoresistance (MR) measurements have been made on sputtered hydrogenated amorphous silicon films for magnetic fields up to 1 T in the temperature range 100-300 K. The effect of a systematic increase of the hydrogen content in the films has been studied. In α-Si films with no hydrogen MR was always negative and it increased with decreasing temperature. In hydrogenated films MR was positive and its magnitude about a decade larger than the negative MR. The positive MR showed a maximum at low fields ∼ 0.2 T. Most of the observed features in the temperature range 200 K < T < 300 K can be explained by a hopping model in which MR is a result of the modification of the spin-flip relaxation time by the external magnetic field. It is concluded that the change of sign in MR with increasing hydrogen content is most probably due merely to the decrease in the density of states in the gap which causes a narrowing in the distribution of the hopping relaxation times. The obtained results support the idea that the variable range hopping mechanism dominates the electrical conduction in sputtered α-Si below room temperature. Further, the results indicate that spin dependent effects have a strong influence on the electrical transport in amorphous semiconductors.

Since the Mössbauer isomer shift is directly connected with the electron charge density inside the nucleus, which in turn is determined by the electronic structure of the solid, it is evident that isomer shift measurements reflect to a certain extent both the band structure of the host crystal and the electronic configuration of the Mössbauer impurity atom. As an example the electronic structure of a tin (¹¹⁹Sn) atom (ion) substitutionally implanted in several semiconducting materials is discussed: whereas in group IV semiconductors tin represents an isovalent impurity giving rise only to one isomer shift line, in A^IIIB^V semiconducting compounds two isomer shift values are observed corresponding to a donor and acceptor configuration.

Relaxation oscillations at subharmonics of the Josephson frequency may occur in a resistively shunted superconducting tunnel junction, where the shunt loop has a large inductance. We have studied the variation in the relaxation period with the shunt inductance and resistance, the Josephson critical current, and the bias current. Steps in the current-voltage curve were noted at bias points where the subharmonic number changed. Radiation powers of the order of a nW were coupled out of small, single, tunnel junctions at 10 GHz. Linewidths scaled with the frequency square as compared with those of the Josephson mode oscillations. Wide signals centered around harmonics of the relaxation frequency were due to amplified noise. Injection locking to a fraction of the frequency of a relatively strong signal occurred, while weaker signals were amplified. Gains up to 15 dB were achieved within a fairly large bandwidth, but the noise temperature was high. The importance of parametrically up-converted noise from low idler frequencies is emphasized.

We have investigated phase-locking and voltage-locking in systems consisting of two planar dc-SQUIDs* located close together in the same superconducting film. The locking strength was found to be strongly dependent on the fluxoid state of both SQUIDs. We present a simple model for the inductive coupling between such two dc-SQUIDs and compare the prediction of the model with the experimental results.

The series of one-dimensional organic conductors (TMTSF)₂X, X = PF₆^-, AsF₆^-, NO₃^-, BF₄^-, ReO₄^-, ClO₄^- and others has shown a remarkable richness of interesting properties. Most members of the series become superconductors under modest pressures ranging from ambient to 15 kbar at temperatures between 0.9 and 1.6 K. Metal-insulator transitions occur in the majority of compounds in the family but the nature hereof and the transition temperature is widely different among the individual members: The hexafluorides seem to be very much alike and they all probably undergo antiferromagnetic phase transitions into spin density wave states near 15 K. By contrast the ReO₄^- compound undergoes a Peierls type of transition near 180 K. (TMTSF)₂ClO₄ does not undergo a complete metal-insulator transition and remains highly conducting down to 1.5 K at which temperature it becomes superconducting. However, an anomaly is found in the spin susceptibility of (TMTSF)₂ClO₄ near 6 K, so the existence of spin density waves in this material is not yet ruled out.

Experimental results reported here include electron spin resonance, nuclear magnetic resonance, microwave conductivity, microwave dielectric constant and static magnetic susceptibility of various members of the family of organic conductors (TMTSF)₂X.

The organic conductors, (TMTSF)₂ PF₆ and (TMTSF)₂ AsF₆ undergo both at ambient pressure a phase transition toward an antiferromagnetically ordered ground state. We review the obtained magnetic susceptibility data and show, that these are almost exactly the same for the two compounds. Both materials have magnetic ordering temperature close to 12 K. In the antiferromagnetically ordered state, spin-flop transitions are seen. The data are discussed on the basis of Overhauser's treatment of itinerant antiferromagnetism.

The instability of the homogeneous state in a ferromagnetic semiconductor is studied. The electronic part of the free energy is determined using Thomas-Fermi statistical model and the magnetic part is calculated by the molecular field approximation including the RKKY-interaction. The inhomogeneity consists of a small magnetically polarized region with a high electron density surrounded by a less polarized positively charged depletion layer. The inhomogeneous state is found to be stable in a relatively broad temperature range around the Curie temperature at low and intermediate doping densities. The stability range shrinks in an applied magnetic field. At fields exceeding about 3 T or at doping densities larger than 10²¹ cm^-3 the inhomogeneous state is no more stable.

We report measurements in the temperature range 10-300 K of the decay time of excitons bound to isolated nitrogen atoms (A-line) and to pairs of nitrogen atoms with different internuclear distances (NN-lines). It is observed that the decay time of the A-line is about 45 ns up to 20 K, drops to about 4 ns at 40 K and then increases rapidly. The miminum at 40 K allows of the estimate 3 × 10^-14 cm² of the cross-section for capture of excitons by NN pair centres. Above 70 K the measured decay curves for the A-line and NN₁-line indicate that thermal equilibrium is established in the system after a few times the decay time of the A-line.

Schottky barriers formed by a conducting polymer, polypyrrole, grown on crystalline and amorphous hydrogenated silicon have been studied. Three techniques have been used to measure the height of the Schottky barrier: I(V), C(V) and internal photoemission. All measurements give a large barrier height. It can be concluded that the work function of polypyrrole, which is a metallike conducting polymer, is high and close to that of gold.

X-ray crystal structure determinations have been carried out on (TMTSF)₂X, where TMTSF = tetramethyltetraselenafulvalene and X = PF₆^-, ClO₄^-, ReO₄^-. These solids exhibit superconductivity at low temperature (∼ 1 K); for PF₆^- and ReO₄^- hydrostatic pressure of the order 10 kbar is also required. This paper deals with a description and a comparison of the crystal structures, which have been determined at ambient temperature and pressure. Possible relationships between structural features and physical properties are discussed.

The single independent component Ā of the internal strain tensor of silicon has been determined by observing the 006 forbidden reflection whilst applying a stress along the [10] axis. A value Ā = -0.18 ± 0.01 has been found. Advantages of the energy dispersive method are indicated.

X-ray diffuse scattering studies at room temperature for seven one-dimensional metallic complexes are reported. The crystal structure and the temperature dependence of both the electrical conductivity and X-ray diffuse scattering for Rb₃(H₃O)_0·2 [Pt(CN)₄] (O₃SO·H·OSO₃)_0·5·0.8H₂O are reported and related to the development of a non-Peierls superstructure below room temperature. The Rb complex is unique in that it contains a polyatomic (> 2) trianion and a ratio of three alkali metal ions per Pt atoms instead of the usual two.

The zero field steps in the current-voltage characteristic of Josephson tunnel junctions are explained as a manifestation of the oscillatory motion of N solitons in the tunnelling barrier. Here N is the step number. The solitons evolve according to a perturbed sine-Gordon equation and numerical solutions of this equation explain the zero field steps as well as many features of the emitted microwave radiation.

Evolution of one dimensional charge density waves in organic conductors can be described by a double sine-Gordon equation. Interaction between solitary wave solutions (2π-kinks and 2π-antikinks) of this equation is examined numerically. Waves of the same kind show a quasi-solitonic behaviour, i.e., the waves survive the collision, and only a minor change in energy and momentum is observed. The result of the interaction between waves of different kind depends on the initial velocity of the interacting waves. For velocities less than a certain threshold velocity the waves enter a breather-like mode which rapidly dies out, while for larger velocities the waves show a quasisolitonic behaviour.

A comparison is made for the bulk modulus B and the shear moduli C and C' of several f.c.c. transition metals and alloys of the 3d series. The moduli of paramagnetic alloys and magnetically ordered alloys are plotted as a function of electron concentration X. A correction is made to account for the differences in the atomic volume in different alloy systems. In addition to band structure effects the magnetic order seems to give important contributions to the composition dependence. Some magnetic effects in the moduli are observable above the ordering temperature in alloys with 8.6 < X < 8.9. The relative effect of magnetism seems to be larger in the modulus C' than in C. The low values of the moduli in the 3d series of transition metals as compared with the other two series also seem to be related to magnetic effects.

The a.c. susceptibility χ' has been measured on amorphous films of composition Gd_0.37Al_0.63 in the frequency range 1.0-2.1 MHz. The results show that the cusp in χ' typical for a spin-glass transition is frequency dependent. A new model is proposed as an extension of the classical Casimir-du Pré model. The whole spin-glass system is represented by several subsystems, each with a different thermodynamic temperature. By varying the measuring frequency, ω, relaxation processes between the different subsystems may be described by time-dependent temperature differences. It is also suggested that slow relaxations between the spins may produce a frequency-dependent cusp as well as a strong field dependence of χ'. In this case it may happen that there is no thermal equilibrium between certain parts of the spin-system. These parts may then relax directly to the lattice. The measured field dependence of χ'(H) for the present system has been analysed using expansions of magnetothermal parameters.

The magnetic susceptibility of a commensurate Cr alloy with 2 at% Si and 0.5 at% Fe has been measured as a function of the temperature in the range 4-340 K in order to check the previously established quadratic dependence on the Fe concentration of a Curie-Weiss part and the linear dependence on the Fe concentration of a temperature independent part of the susceptibility below the Néel temperature for commensurate CrFe and CrSiFe alloys with less than 4 at% Fe. These facts are verified by the new results although with slight modifications because of disturbing interactions between the Si and the Fe atoms. The Curie-Weiss part is ascribed to nearest neighbour pairs of Fe atoms and the temperature independent part to isolated Fe atoms.

Electron localization and delocalization effects in the MNN Auger spectra of 4d transition metals have been studied. Based upon Auger peak energies, half widths and electron binding energies a semi-empirical model, similar to a model proposed by Lin Yin et al. [2] for 3d metals, has been used to characterize the M₄₅N₂₃N₂₃ and M₄₅N₄₅N₄₅ transitions. A basic parameter, δΔ, can be written as a difference between a holehole interaction and a relaxation term. The relaxation term expresses the net difference in relaxation between single- and double-hole final states. Large values of Δ mean large changes in the valence electron energy due to ineffective screening, a high degree of localization of electrons, and a quasi-atomic behaviour is expected.

Quantitative estimates for the delocalization of the 4d electrons and the relaxation terms have been carried out. Analysis of measured Auger spectra of yttrium, molybdenum and silver has been combined with values deduced from published Auger spectra of the other elements of the second transition metal group. A clear trend through the group is observed. For the M₄₅N₂₃N₂₃ transitions, involving core electrons, an almost constant relaxation is found indicating a high degree of electron localization. However, for the M₄₅N₄₅N₄₅ transition, involving valence electrons, increase in the relaxation term with increasing charge number is found. This behaviour is interpreted as a high degree of electron delocalization in the beginning of the group and a quasi-atomic behaviour in the case of Ag.

Without abstract; see full text