No abstract

Crystal growth, viscosity, and melting were studied in Ge2Sb2Se5 bulk samples. The crystals formed a compact layer on the surface of the sample and then continued to grow from the surface to the central part of the sample. The formed crystalline layer grew linearly with time, which suggests that the crystal growth is controlled by liquid-crystal interface kinetics. Combining the growth data with the measured viscosities and melting data, crystal growth could be described on the basis of standard crystal growth models. The screw dislocation growth model seems to be operative in describing the temperature dependence of the crystal growth rate in the studied material in a wide temperature range. A detailed discussion on the relation between the kinetic coefficient of crystal growth and viscosity (ukin ∝ η(-ξ)) is presented. The activation energy of crystal growth was found to be higher than the activation energy of crystallization obtained from differential scanning calorimetry, which covers the whole nucleation-growth process. This difference is considered and explained under the experimental conditions.

Chalcogenide glass-formers are being used in a remarkable range of various optoelectronic, photonics, photoconducting, sensing and memory device applications. The knowledge of viscosity is essential for the processing of any glass-forming material, in particular for the fabrication of precise optical elements, which is the main application field of chalcogenide glasses. This work presents an extensive collection of all available viscosity data for chalcogenides, including the measurement methods. The Mauro-Yue-Ellison-Gupta-Allan (MYEGA), Arrhenius and VFT equations are used to fit the temperature dependences of viscosity. The viscosity glass transition temperatures, fragilities and apparent activation energies are calculated from these fits. Consequently, these parameters are discussed with regard to the compositional evolution of the respective chalcogenide systems.

Crystal growth rates in GeSbSe bulk glass and thin film were measured using optical and scanning electron microscopy under isothermal conditions. The studied temperature region was 255-346 °C and 254-286 °C for bulk glass and thin film, respectively. The compact crystalline layer growing from the surface into the amorphous core was formed in bulk glasses and no bulk crystallization was observed. In the case of thin films, needle-shape crystals were formed. The crystalline layer and needle-shape crystals grew linearly with time that corresponds to a crystal growth controlled by the crystal-liquid interface kinetics. In the narrow temperature range, crystal growth rates exhibit simple exponential behavior, so the activation energies of crystal growth for the studied temperature regions were estimated (E = 294 ± 6 kJ/mol for bulk glass and E = 224 ± 12 kJ/mol for thin film). Viscosity of GeSbSe material was measured in the region of the undercooled melt and glass. The extrapolation of viscosity data into the immeasurable, but important, temperature range is discussed. The experimental growth data were combined with melting and viscosity data and the appropriate growth models were proposed to describe crystal growth in a wide temperature region. The standard crystal growth models are based on a simple proportionality of the crystal growth rate to the viscosity (u ∝ η). This simple proportionality holds for the bulk material. Nevertheless, in the thin films the decoupling of the crystal growth rate from the inverse viscosity occurs, and the standard kinetic growth models need to be corrected. Such corrections provide better description of experimental data and more realistic value of the parameter describing the mean interatomic distance in the crystal-liquid interface layer, where the crystal growth takes place.

The crystal growth velocity of spherulitic As2Se3 in a supercooled melt of the same composition was studied by optical microscopy and thermoanalytical methods in isothermal and nonisothermal conditions. The time dependence of crystal size is linear, which suggests the crystal growth is controlled by interface kinetics. Crystal growth velocity was determined as the slope of these linear dependences. The experimental results presented in this paper considerably extend the previously reported range of crystal growth velocity. All isothermal crystal growth velocity data can be well described by the standard two-dimensional surface nucleated growth model (2Dsg) including crystal growth viscosity decoupling (ξ = 0.647). The activation energy of crystal growth for microscopic experiments is in a good agreement with values obtained from thermoanalytical experiments, and the ratio of the activation energy of crystal growth and the activation energy of viscous flow well corresponds to an independently determined decoupling parameter. The same model successfully describes also crystalline layer thickness and growth pattern at the amorphous As2Se3 surface in nonisothermal conditions.