Amorphous Phase Change Materials: Structure, Stability and Relation with Their Crystalline Phase

Abstract

Phase Change Materials should be stable enough in their amorphous phase to achieve a durable data retention, however they should also be bad glass formers to be able to recrystallise at high speed. To understand these contradicting properties, we construct models of amorphous Ge-Sb-Te systems using Ab Initio Molecular Dynamics and analyse the structures in relation with the relevant crystalline state. We show that structural patterns that are precursors of the crystalline phase exist in the amorphous state and we identify the signature of the various types of local atomic orders in the X-ray absorption spectra that we compute using Density Functional Theory. We first analyse the mechanical properties of the amorphous phase in the framework of the Maxwell rigidity theory, showing that all efficient Phase Change Materials deviate from the perfect glass and are mechanically stressed-rigid. Additionally, we show that the stability of Phase Change Materials is related to the density of low frequency vibrational modes (Boson peak). We describe how an adequate doping can result in an increased stability of the amorphous phase while keeping intact the phase change ability of the material.

Nanoscale Mg–B via Surfactant Ball Milling of MgB2: Morphology, Composition, and Improved Hydrogen Storage Properties

“…The XAS simulations were carried out in VASP using an implementation that computes the required matrix elements within the PAW formalism, including a self-consistent treatment of the core−hole effects on the unoccupied states. 38 Theoretical spectra were broadened with a 0.5 eV Gaussian. In this study, DFT calculations were performed on a slab model consisting of 8-layers of MgB 2 (001), alternating between Mg and B, separated by 19.9 Å of vacuum with boron exposed on one side and magnesium exposed on the other.…”

“…Forces were relaxed to below 0.005 eV/Å. The XAS simulations were carried out in VASP using an implementation that computes the required matrix elements within the PAW formalism, including a self-consistent treatment of the core–hole effects on the unoccupied states . Theoretical spectra were broadened with a 0.5 eV Gaussian.…”

The interplay between Peierls distortions and metavalent bonding in IV–VI compounds: comparing GeTe with related monochalcogenides

“…The heaviest compounds have steeper effective repulsive potentials between cores which prevent any distortion, and are therefore cubic (PbSe, PbTe). This view of the chemical bonding in IV-VI chalcogenides has been widely used to account for many phenomena observed in these compounds, for instance to account for the temperature and pressure induced phase transitions (the distorted systems become cubic [24,25]) or for the characteristics of the liquid and amorphous phases [17,[26][27][28][29]. However, we are not aware of any theory based on the Peierls distortion or covalent bonding, which can predict for which chalcogenides a pronounced difference exists between the amorphous and crystalline state.…”