Crystal growth in Se70Te30 thin films followed by SEM and in situ XRD

Abstract

The isothermal crystal growth kinetics in Se70Te30 thin films was investigated using the microscopy and in situ X-ray diffraction (XRD) measurements. Plate-like crystals grew linearly with time which is the sign of liquid-crystal interface kinetics. In the studied temperature range, from 68 °C to 88 °C, crystal growth rates exhibit simple exponential behavior with an activation energy of crystal growth EG = 168 ± 12 kJ mol−1. The growth data obtained from the microscopy measurements were combined with viscosity data, melting parameters and the appropriate crystal growth model was assessed. The relation between the kinetic coefficient of crystal growth and viscosity (u∝η-ξ) is described in detail, and a correction of the standard growth model is suggested. The crystal growth data obtained from the in situ XRD measurements were described using the Johnson-Mehl-Avrami nucleation-growth model with the Avrami exponent m = 2.2 ± 0.2. The activation energy of the overall crystallization process EA was estimated and its value is 171 ± 11 kJ mol−1.

Extended Study on Crystal Growth and Viscosity in Ge–Sb–Se Bulk Glasses and Thin Films

“…It was shown that this assumption is not valid below 1.2 T g where the crystal growth rate and viscosity can decouple. It was also shown in our previous published work that the standard growth models do not describe the experimental growth data well if the decoupling of the crystal growth rate and viscosity occurs, and the correction of standard models is necessary. Ediger proposed a simple way to test the decoupling which is based on a power law dependence of u kin on the viscosity: where u kin is a kinetic part of crystal growth rate ( u kin = u /[1 – exp­(−Δ G / RT )]) and the exponent ξ < 1 expresses the extent of decoupling between crystal growth rate and viscosity.…”

“…Standard crystal growth models assume a simple proportionality of crystal growth rate to inverse viscosity ( u ∝ η –1 ), , according to the standard Stokes–Einstein equation, which is frequently used to replace the effective diffusion coefficient by the viscosity. Nevertheless, during the past two decades the relation between viscosity and crystal growth rate was investigated and tested in different types of glasses. ,− This analysis was found to be important because some fragile glasses show a decoupling of the crystal growth rate from the inverse viscosity, and the relation between them is corrected for the parameter ξ ( u ∝ η –ξ ). The growth models can be then corrected to describe the growth data in the measured temperature range. , …”

“…Nevertheless, during the past two decades the relation between viscosity and crystal growth rate was investigated and tested in different types of glasses. ,− This analysis was found to be important because some fragile glasses show a decoupling of the crystal growth rate from the inverse viscosity, and the relation between them is corrected for the parameter ξ ( u ∝ η –ξ ). The growth models can be then corrected to describe the growth data in the measured temperature range. , …”

“…The growth models can be then corrected to describe the growth data in the measured temperature range. 24,25 The aim of this work is direct investigation of the crystal growth kinetics in Ge 18 Sb 28 Se 54 bulk glasses and thin films using optical and scanning electron microscopy. The extended study of Ge 18 Sb 28 Se 54 system is performed including the study of melting parameters and temperature dependence of viscosity for the bulk material.…”

Analysis of crystal growth and viscosity in Ge-Sb-Se-Te undercooled melts

“…The standard growth models are based on an assumption, that the own transport of the structural units to the liquid-crystal interface is driven by self-diffusion, which can be substituted by inverse viscosity according to the Stokes-Einstein-Eyring relation [49]. Nevertheless, in recent studies on crystal growth [13,37,38,48,[50][51][52][53][54] a significant decoupling of crystal growth rate and viscosity occurred, therefore, the standard crystal growth models were corrected to the decoupling phenomenon by including the decoupling parameter  into the standard crystal growth models (Eq. 5).…”