Field-dependent negative-U model and switching in chalcogenide glasses

Using the Anderson negative-U model, with harmonically bound ionic charges, one finds that the electric dipoles of singly (D₀) and doubly (D₋) occupied dangling bonds can be streched by a strong electric field. The D₀ and D₋ energies become field-dependent, with the splitting E₀(F₋)−E_(F) vanishing at a critical field F_c when D₀ becomes more stable. Assuming that Zener tunneling to extended states occurs only from singly occupied states, but not from D₋ states, a qualitative picture for Ovshinsky switching follows: there will be a jump in the D₀ occupation and hence in the carrier concentration at F = F_c. Field magnitudes and delay times are estimated. Memory switching is also discussed.     

Diffusion in bulk-metallic glass-forming Pd–Cu–Ni–P alloys: From the glass to the equilibrium melt

Since the discovery of bulk-metallic glasses there has been considerable research effort on these systems, in particular with respect to mass transport. Now the undercooled melt between the melting temperature and the caloric glass transition temperature, which has not been accessible before due to the rapid onset of crystallization, can be investigated and theories can be tested. Here we report on radiotracer diffusion measurements in metallic bulk-glass-forming Pd–Cu–Ni–P alloys. Serial sectioning was performed by grinding and ion-beam sputtering. The time, temperature as well as the mass dependence, expressed in terms of the isotope effect E, of Co-diffusion were investigated. In the glassy state as well as in the deeply supercooled state below the critical temperature T_c, where the mode coupling theory predicts a freezing-in of liquid-like motion, the experimentally determined very small isotope effects indicate a highly collective hopping mechanism involving some ten atoms. Below T_c the temperature dependence shows Arrhenius-type behavior with an effective activation enthalpy of 3.2 eV. Above T_c the onset of liquid-like motion is evidenced by a gradual drop of the effective activation energy and by the validity of the Stokes–Einstein equation, which is found to break down below T_c. This strongly supports the mode coupling scenario. The Stokes–Einstein equation is presently tested for other constituents of the alloy. The Co isotope effect measurements, which have never been carried out near T_c in any material, show atomic transport up to the equilibrium melt to be far away from the hydrodynamic regime of uncorrelated binary collisions.     

The role of stress development and relaxation on crystal growth in glass

The role of internal stresses energy is usually neglected when crystallization kinetics is considered. The common argument is that stress is relaxing too fast to affect the process. In this article we develop a generalized formalism to describe steady-state growth kinetics in viscoelastic media. The residual stress energy results from interplay between the rate of stress development (due to the propagation of a crystal throughout the matrix) and rate of stress dissipation (due to relaxation of the viscous matrix). The degree to which the stress energy can relax depends on the ratio of τ_c/τ_r the characteristic time for crystal growth, τ_c, and the relaxation time, τ_r. The present results challenge the widespread fallacy that internal stresses relax too fast to affect crystal growth. Our model explains the often observed lack of agreement between the theoretical predictions (without taking into account the stresses energy development) and experimental data.     

The dynamics of wet granular matter

We present a numerical study of a shear-induced solid–fluid transition in wet granular matter in order to show the self-organized critical behavior close to the transition point. The continuous time simulation is based on a simple model considering both the cohesive forces induced by the adsorbed liquid amount and the repulsive forces due to the hard core interaction of the granules. Dissipation is assumed to be entirely due to the hysteretic character of the cohesive forces. The aim of our analysis concerns the crossover from a solid like behavior to a mobile ergodic state under the influence of an external force field F exceeding the critical force F_c. Diffusion coefficients, dissipation and kinetic order parameters can be expressed as characteristic scaling laws.     

Ordering of tysonite structure in an as-grown Er0.715Ca0.285F2.715 crystal and in a component of annealed two-phase crystal of the Er0.67Ca0.33F2.67 composition

The structures of the as-grown tysonite phase (TP) Er_0.715Ca_0.285F_2.715 as grown and the tysonite component in a crystal of the Er0.67Ca0.33F2.67 composition annealed at 760°C have been determined by X-ray diffraction methods from the main reflections in the sp. gr. P6₃/mmc, Z = 2. Ca²⁺ cations in the Er_0.715Ca_0.285F_2.715 structure occupy the 2c site on the 6₃ axis, thus confirming the conclusion about the symmetrizing effect of alkaline earth cation, which was made previously for the Y_0.715Ca_0.285F_2.715 compound. Er³⁺ cations are disordered over 6h sites around this axis. The Er_0.67Ca_0.33F_2.67 composition contains Er_0.67+δCa_0.33?δF_2.67+δ TP and inclusions of distorted fluorite phase rhβ-Ca₈Er₅F₃₁. The tysonite and two fluorite lattices have no “correct” relationship. Ca²⁺ cations and a half of TP Er³⁺ cations are fixed in the 2c sites, while the other Er³⁺ cations are disordered over the 6h sites. Superstructural ordering occurs in Er_0.715Ca_0.285F_2.715 crystals; however, the process is not completed, as follows from the character of the diffraction pattern. Weak satellite reflections in the diffraction pattern of the Er_0.67Ca_0.33F_2.67 composition are indicative of the development of TP ordering processes but remain unfit for structural calculations.     

Kinetics of crystallization of titanium based binary and ternary amorphous alloys

Metallic glasses are kinetically metastable materials. These amorphous materials can be transformed into a crystalline state by both isothermal and isochronal methods. The study of this transformation, and hence the thermal stability of metallic glasses, are important from an application view-point. In the present work, the non-isothermal crystallization kinetics of two titanium-based amorphous alloys namely, Cu₅₀Ti₅₀ and Ti₅₀Ni₃₀Cu₂₀, are reported. The activation energies for crystallization, E_c for both the systems have been evaluated using different non-isothermal methods viz. derived through Kissinger, Augis and Bennet and Ozawa. The values of E_c obtained using these methods are consistent for both the metallic glasses and it is found that E_c for the ternary metallic glass is considerably higher than the binary metallic glass. The increase in the activation energy on the substitution of Ni in the Cu–Ti metallic glass suggests the increase in the thermal stability.     

Switching phenomena and dielectric breakdown in amorphous Nb2O5 films

Repititive switching from the high-to the low-resistance states in thin amorphous Nb₂O₅ films are investigated. The switching processes are found to be associated with four conduction states. The transitions among the first three states are reversible, exhibiting a cyclic switching behavior. The resistive transitions are abrupt, taking place within a range of less than a millivolt. The switching time from the first conduction state to others is <2 ns, while the switching from the second state occurs with a voltage-dependent delay time. The values of resistivity, thermal activation energy and I–V characteristics of the four conduction states are given and show that the final or irreversible dielectric breakdown process occurs throuh the formation of highly conductive filaments. The current-voltage characteristics for the four conduction states are briefly discussed.     

X-ray and neutron diffraction study of the defect crystal structure of the as-grown nonstoichiometric phase Y0.715Ca0.285F2.715

This work begins a series of papers aimed at studying the defect structure of nonstoichiometric phases R _{1 ? y} Ca _y F_{3 ? y} with a tysonite-type (LaF₃) structure. In the single-crystal structure of Y_0.715Ca_0.285F_2.715 with a tysonite-type small unit cell (sp. gr. P6₃/mmc, a = 3.9095(2) Å, c = 6.9829(2) Å; Z = 2; R _w = 2.16%), the displacements of Y³⁺ cations and F^2? anions from 6₃ symmetry axes were observed for the first time. The X-ray diffraction pattern shows weak satellites insufficient for structural calculations. The LaF₃ structure type is stabilized up and down on the temperature scale due to anion vacancies and the symmetrizing effect of Ca²⁺ cations lying on 6₃ symmetry axes. At 120°C the fluoride-ion conductivity in the nonstoichiometric phase Y_0.715Ca_0.285F_2.715 is five orders of magnitude higher than that in the stoichiometric phase β-YF₃. The transition to a superionic state is caused by a deviation from stoichiometry and is not associated with reconstructive phase transformation.     

Effect of devitrification on ion motion in lithium-disilicate glass

Li-disilicate glass-ceramics consist of microcrystallites imbedded in the glassy Li₂O · 2SiO₂ matrix where the number and size of the crystallites depend on the devitrification heat treatment. To assess ion motion in these model glass-ceramics, we have measured the temperature dependence of the dc conductivity, σ_dc, and the ⁷Li nuclear spin relaxation (NSR) rate, 1/T₁, in samples with various crystalline fraction, c, ranging from c = 0 (pure glass) to c = 1 (fully devitrified polycrystalline ceramic). The Cole–Cole presentation of the complex impedance shows two separate arcs caused by the remarkable difference of the ionic motion in the glassy and crystalline phase. These two arcs correspond to a bi-exponential decay of the ⁷Li nuclear spin magnetization where the resulting two NSR rates are induced by the ionic motion in the two phases. Thus the NSR and σ_dc data provide a comprehensive picture of the ionic motion in the glassy and crystalline phases. In particular, the ionic motion is the fastest in the glass; then at lower values of c we observe a metastable crystalline phase with ionic motion much greater than in the stable (LS₂) crystalline phase existing at large c-values.     

Magnetic properties of liquid 3d transition metal–Si alloys

The magnetic susceptibility has been measured for liquid transition metal–Si alloys (TM_1−cSi_c, TM = Ni, Co, Fe, Mn) as a function of temperature. The magnetic susceptibilities of liquid Ni_1−cSi_c with c ⩾ 0.3, liquid Co_1−cSi_c with c ⩾ 0.4 and liquid Fe_1−cSi_c with c ⩾ 0.6 were found to be almost independent of temperature, which suggests that the Ni, Co and Fe ions on the Si side are in the non-magnetic state. Liquid Ni_1−cSi_c with c ⩽ 0.2, liquid Co_1−cSi_c with c ⩽ 0.3 and liquid Fe_1−cSi_c with c ⩽ 0.5 show the Curie–Weiss behavior with a reasonable value of effective Bohr magnetic number. Liquid Mn_1−cSi_c with c ⩽ 0.2 and c ⩾ 0.6 were found to be in their non-magnetic state. However, the Curie–Weiss behavior was observed for liquid Mn_1−cSi_c with c = 0.3, 0.4 and 0.5. The magnetic susceptibility of their liquid alloys in the non-magnetic state has been studied by using Anderson model. The density of 3d states at the Fermi level can be estimated from the data of magnetic susceptibility.     

Crystallochemical model of ion-transport percolation in nonstoichiometric M 1−x RxF2+x phases with the defect CaF2-type structure

A crystallochemical model of ion-transport percolation in M _1?x R_xF_2+x (M = Ca, Sr, Ba; R = RE) solid solutions with the “defect” CaF₂-type structure has been suggested. Within this model, the percolation thresholds in M _1?x R_xF_2+x crystals with tetrahedral R ₄F_26? and octahedral R ₆F_37? clusters are considered, whose existence is highly probable in these disordered fluorite phases. It is established that the calculated percolation thresholds x _c = 2.8 mol % for nnn (next nearest neighbors) [(R ₆F₃₇) _M6F32 ?F_i] and x _c = 4.7 mol % for nn (nearest neighbor) [(R ₄F₂₆) _M6F32 ?F_i] are in satisfactory accord with the experimental percolation thresholds determined from the conductometric data.     

Some features relevant to switching processes in the amorphous semiconductor Si12Ge10As30Te48

The current-voltage characteristics of the amorphous semiconductor glass Si₁₂Ge₁₀As₃₀Te₄₈ have been measured under various experimenal conditions. The experimental results show that the resistance of the device in the ‘off’ state and the threshold voltage for the onset of switching action decrease with increasing the maximum current (I_p) passing through the device in the ‘on’ state, and that the threshold input power to set in the switching action is practically independent of temperature. The stability and the consistency of the device depend on the magnitude of I_p. When I_p is increased to a certain value the glass within and near the current filament between the electrodes become softened, and when it reaches a critical value the device is changed from its ‘switching-on’ state to a ‘memory’ state. All the results are in good agreement with the model that the filament formed to cause switching consists of a mixture of crystalline domains and amorphous domains with phase separations.     

Study of correlation between morphological structures and composition inhomogeneity for Ca0.9Ho0.1F2.1 crystals

Optical methods and local X-ray spectroscopic analysis were used to study the correlation between the morphological structures and the composition inhomogeneity for Ca_0.9Ho_0.1F_2.1 crystals. It is found that, at two nearby points, the characteristics of uniformity of the Ca and Ho distribution (the homogeneity criterion S _c/2σ, the coefficient of variation S _c/N?, and the level of homogeneity W ₉/C) may differ by more than two orders of magnitude if these points belong to morphologically different regions with different prehistory.     

Effect of impurities on the metastable zone width of solute–solvent systems

The novel approach to interpret the metastable zone width obtained by the polythermal method using the classical theory of three-dimensional nucleation proposed recently [K. Sangwal, Cryst. Growth Des. 9 (2009) 942] is extended to describe the metastable zone width of solute–solvent systems in the presence of impurities. It is considered that impurity particles present in the solution can change the nucleation rate J by affecting both the kinetic factor A and the term B related with the solute–solvent interfacial energy γ. An expression relating metastable zone width, as defined by the maximum supercooling ΔT_max of a solution saturated at temperature T₀, with cooling rate R is proposed in the form: (T₀/ΔT_max)²=F(1−Z ln R), where F and Z are constants. The above relation can also be applied to describe the experimental data on maximum supercooling ΔT_max obtained at a given constant R as a function of impurity concentration c_i by the polythermal method and on maximum supersaturation σ_max as a function of impurity concentration c_i by the isothermal method. Experimental data on ΔT_max obtained as a function of cooling rate R for solutions containing various concentrations c_i of different impurities and as a function of concentration c_i of impurities at constant R by the polythermal method and on σ_max as a function of impurity concentration c_i by the isothermal method are analyzed satisfactorily using the above approach. The experimental data are also analyzed using the expression of the self-consistent Nývlt-like approach [K. Sangwal, Cryst. Res. Technol. 44 (2009) 231]: ln(ΔT_max/T₀)=Φ+β ln R, where Φ and β are constants. It was found that the trends of the dependences of Φ and β on impurity concentration c_i are similar to those observed in the trends of the dependences of constants F and Z on c_i predicted by the approach based on the classical nucleation theory.     

Load relaxation of a flat rigid circular indenter on a gel half space

When a saturated elastic gel is suddenly indented to a fixed displacement, δ, by a flat circular rigid indenter, the solvent cannot escape immediately, this gives rise to a pressure gradient in the solvent and the solvent flows until the pressure in it goes to zero everywhere, and all the stresses are transferred to the elastic network. This load transfer process causes, the indenter load, F, to decrease with time t, with a characteristic time constant, τ, which is given by a²/D_c, where D_c is the cooperative diffusion coefficient and a is the radius of the circular punch. This load relaxation behavior can be used to extract the elastic moduli and the cooperative diffusion coefficient of the gel. A detailed finite element analysis shows the load relaxation as a function of normalized time (t/τ) and how it depends on the shear modulus G and the Poisson’s ratio ν of the drained gel. Using this result, we found that the energy release rate available for detaching the punch in a pull-off test can be estimated. It is found that the energy release rate decreases with time for a fixed displacement.     

X-ray diffraction study of 2,2′,4,4′,6,8,8′,10,10′,12-decamethylbicyclo[5.5.1]hexasiloxane and 2,2′,4,4′,6,8,8′,10,10′,12-decamethylbicyclo[5.5.1]-9-carbahexasiloxane

The crystal structures of two organosilicon compounds are studied by X-ray diffraction at 200 K. Crystals of 2,2′,4,4′,6,8,8′,10,10′,12-decamethylbicyclo[5.5.1]hexasiloxane (C₁₀H₃₀O₇Si₆) are monoclinic, a = 12.260(1) Å, b = 11.716(1) Å, c = 32.169(2) Å, β = 92.802(2)°, space group C2/c, Z = 8, wR2 = 0.132, and R1 = 0.053 for 3989 reflections with F > 4σ(F). Crystals of 2,2′,4,4′,6,8,8′,10,10′, 12-decamethylbicyclo[5.5.1]-9-carbahexasiloxane (C₁₁H₃₂O₆Si₆) are triclinic, a = 10.709(1) Å, b = 11.046(1) Å, c = 12.111(1) Å, α = 117.059(1)°, β = 95.566(1)°, γ = 108.088(1)°, space group $P\bar 1$ , Z = 2, wR2 = 0.115, and R1 = 0.048 for 3948 reflections with F > 4σ(F). It is found that the substitution of the methylene group for the oxygen atom is accompanied by a minor conformational change in the bicyclic fragment and the SiCSi angle changes from the tetrahedral angle to 121.0(1)°. The high variability of the SiOSi bond angles is confirmed.     

Electron effects of the substituents on the structure of 4,4′-dipyridylium dications

The crystal structure of N,N′-di(2-hydroxyethyl)-4,4′-dipyridylium diperchlorate C₁₄H₁₈Cl₂N₂O₁₀ is determined by X-ray diffraction at 293 K. The crystals are monoclinic, a = 6.446(1) Å, b = 14.479(3) Å, c = 9.912(1) Å, β = 99.24(2)°, space group P2₁/c, and Z = 2; 1562 reflections measured; and wR2 = 0.086 and R1 = 0.033 for 1209 reflections with F > 4σ(F). It is found that, in the absence of charge transfer from the counterion to the dication, the planarity of the dipyridylium skeleton of the molecule is primarily due to the conjugation between the lone electron pairs of the hydroxyl groups and the π system of the dication. This inference is supported by the intramolecular O?N contact (2.847 Å).     

New insight into phase change memories

The switching mechanism in phase change memories was described on the basis of minimum switching unit: the commuton. A commuton is a minimum cluster of atoms that supports a reversible phase change from high to low electrical conduction state and back under the influence of an external signal. The switching process in a phase change chalcogenide film was modeled using two dimensional cellular automata approach. A system of 50 × 50 cells, each cell containing a commuton, was simulated. In the particular case of Ge₂Sb₂Te₅ (investigated here) this system corresponds to a 30 × 30 nm area. The formation of the percolation path as a function of phase change induced in commutons explains the switching phenomenon. The influence of the percent of defects in the material on the percolation threshold has been studied.     

Electrical properties of a-Ge-Se-In thin films

Steady state and transient photoconductivity has been measured on Ge₂₀Se_80−xIn_x (x = 0, 5, 10, 15, 20) vacuum evaporated thin films. Study of temperature dependent dark conductivity σ_d and photoconductivity σ_ph measurements in the temperature range 303-375 K, shows that the conduction in this glass is through an activated process having single activation energy. The activation energy value of photoconduction is smaller in comparison to activation energy in dark. The photosensitivity shows a maximum value at 10 at.% of In concentration. This is attributed to the decrease in the density of defect states of Ge-Se alloy with increase of In content. The results of intensity dependent steady state photoconductivity σ_ph follow a power law with intensity (F), i.e. σ_ph α F^γ where the value of power γ lies between 0.5 and 1.0, suggesting bimolecular recombination. Rise and decay of photocurrent for different concentration of In shows that photocurrent rises monotonically to the steady state value and the decay of photocurrent is also very fast. An attempt has been made to explain the results on the basis of defects and density of states.     

The iso-structural viscosity,configurational entropy and fragility of oxide liquids

This paper describes how the fragility of a liquid is linked to the ratio between the energy barrier (E_eq) for the equilibrium viscous behavior and that (E_iso) for the non-equilibrium iso-structural viscous behavior. Using the concept of iso-structural viscosity, two functions describing the variation of the configurational entropy (S_c) with temperature (T) are obtained from the Avramov-Milchev (AM) and the Vogel-Fulcher-Tammann (VFT) viscosity equations, respectively. The two S_c(T) functions exhibit different relations to the liquid fragility. The AM S_c(T) function is a power function with the exponent of F ? 1, where F is the AM fragility index. In this case, S_c vanishes at T = 0 K. For the VFT function, S_c vanishes as T is lowered to a finite temperature T₀, whereas it reaches the maximum value S_c,max at infinitively high T. S_c,max is proportional to the VFT fragility index. Thus, the VFT equation is not only a dynamical, but also a thermodynamic model. It is proved that for oxide liquids, the VFT equation describes viscosity data better than the AM equation, provided the pre-exponential factor η₀ is fixed to a generally accepted value, e.g., 10^?3.5 Pa s.