Glass transition of the phase change material AIST and its impact on crystallization

Julian Pries,Julia Charlotte Sehringer,Shuai Wei,Pierre Lucas,Matthias Wuttig

doi:10.1016/j.mssp.2021.105990

Julian Pries, Julia Charlotte Sehringer + Show 3 more

Open Access

https://doi.org/10.1016/j.mssp.2021.105990

Copy DOI

Abstract

Engineering phase change materials (PCM) to realize superior data storage devices requires a detailed understanding of crystallization kinetics and its temperature dependence. The temperature dependence of crystallization differs distinctly between crystallizing from the glassy phase and the undercooled liquid (UCL). Hence, knowing the phase from which crystallization occurs is necessary for predicting the switching ability. Here, we measure the glassy dynamics and crystallization kinetics using calorimetry for heating rates spanning over six orders of magnitude. Our results show that the prominent PCM (Ag,In)-doped Sb2Te (AIST) exhibits a change from crystallizing from the glassy phase to crystallizing from the UCL at a critical heating rate of 5000 K/s. Above the glass transition, the activation energy of crystallization changes drastically enabling rapid crystallization at elevated temperatures.

Highlights

Phase change materials (PCMs) are characterized by their ability to rapidly switch between the two solid states, namely between the amorphous and crystalline phase [1,2,3,4]
Our results show that the prominent PCM (Ag,In)-doped Sb2Te (AIST) exhibits a change from crystallizing from the glassy phase to crystallizing from the undercooled liquid (UCL) at a critical heating rate of 5,000 K/s
Our results show that the glass transition temperature Tg of AIST must be larger than the pre-annealing temperature of 115 °C, which raises the question what the actual glass transition temperature is

Summary

Introduction

Phase change materials (PCMs) are characterized by their ability to rapidly switch between the two solid states, namely between the amorphous and crystalline phase [1,2,3,4]. This technique supports the conclusion from (F)DSC that AIST crystallizes from the glassy phases at low heating rates. If AIST was crystallizing from the UCL, preannealing could lead to higher Tp values as observed in Figure 1 and 3, if pre-annealing decreased the FEM variance.

Results

Conclusion