Omar Abou El Kheir¹, Luigi Bonati², Michele Parrinello², Marco Bernasconi^1*

1 Department of Materials Science, University of Milano-Bicocca, Via R. Cozzi 55, Milano, I-20125, Italy.

2 Atomistic Simulations, Italian Institute of Technology, Via E. Melen 83, Genova, I-16152, Italy.

* Corresponding authors emails: marco.bernasconi@unimib.it

DOI10.24435/materialscloud:a8-45 [version v1]

Publication date: Feb 12, 2024

How to cite this record

Omar Abou El Kheir, Luigi Bonati, Michele Parrinello, Marco Bernasconi, Unraveling the crystallization kinetics of the Ge₂Sb₂Te₅ phase change compound with a machine-learned interatomic potential, Materials Cloud Archive 2024.27 (2024), https://doi.org/10.24435/materialscloud:a8-45

Description

The phase change compound Ge₂Sb₂Te₅ (GST225) is exploited in advanced non-volatile electronic memories and in neuromorphic devices which both rely on a fast and reversible transition between the crystalline and amorphous phases induced by Joule heating. The crystallization kinetics of GST225 is a key functional feature for the operation of these devices. We report here on the development of a machine-learned interatomic potential for GST225 that allowed us to perform large scale molecular dynamics simulations (over 10000 atoms for over 100 ns) to uncover the details of the crystallization kinetics in a wide range of temperatures of interest for the programming of the devices. The potential is obtained by fitting with a deep neural network (NN) scheme a large quantum-mechanical database generated within Density Functional Theory. The availability of a highly efficient and yet highly accurate NN potential opens the possibility to simulate phase change materials at the length and time scales of the real devices.

Materials Cloud sections using this data

Files

File name	Size	Description
NN_potential.zip MD5md5:567a33cd8189751fd47d908cc1969f10	7.9 MiB	DeepMD potential for Ge₂Sb₂Te₅
DB.zip MD5md5:d631b0621c47ceb5802f831069166842	667.4 MiB	The DFT database used to fit the NN potential divided into cubic, hexagonal, and disordered configurations (i.e. liquid, supercooled liquid, and amorphous phases). The disordered configurations are furtherly divided into configurations from DFT MD simulations, from NN MD simulations, and NN metadynamics simulations. The database is provided in DeepMD format.
Crystallization_trajectories.zip MD5md5:8b7b813879efc2749e2ffef45a4defb3	293.5 MiB	Trajectories files of the homogeneous crystallization at 600 K and heterogeneous crystallization at 610 K
crystallization_610K.mp4 MD5md5:2e46bd32d6f83a4849650dfa6a27ef3c	16.3 MiB	Video of heterogeneous crystallization dynamics at 610 K

License

Files and data are licensed under the terms of the following license: Creative Commons Attribution 4.0 International. except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.
Metadata, except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.

External references

Keywords

Neural Network Potential Crystallization Phase Change Memories Ge2Sb2Te5 Phase Change Materials Molecular Dynamics Simulation