Statistical learning for phase-change memory materials

Add to your list(s) Download to your calendar using vCal

• Felix Cosmin Mocanu
• Monday 17 August 2020, 17:00-17:30
• virtual ZOOM meeting ID: 263 591 6003, https://zoom.us/j/2635916003.

If you have a question about this talk, please contact Bingqing Cheng .

The marriage of accurate quantum mechanical simulations and expressive descriptors of local atomic environments has proven extremely fruitful for the predictive modelling of materials. We present several examples which leverage accurate quantum mechanical simulations and statistical learning techniques for the investigation of phase-change memory materials in the ternary Ge-Sb-Te system. The talk will go over: (i) the physics, chemistry and engineering of phase-change memory materials, and their potential uses in beyond-silicon hardware for AI applications, (ii) the use of atomic descriptors for visualising atomic local and global structural similarity in disordered phase-change memory materials, (iii) the training of an approximate interatomic potential for the canonical Ge2Sb2Te5 composition and the Ge-Sb-Te system more generally, from quantum mechanical data (density-functional theory calculations) using the sparse and regularised gaussian approximation potential (GAP) framework [1]. We end by highlighting some of the modelling done with the Ge-Sb-Te GAP potential, the limitations in terms of accuracy and transferability and the scope for future improvement and exploitation of the models.

[1] Felix C. Mocanu et al. “Modeling the Phase-Change Memory Material, Ge2Sb2Te5, with a Machine-Learned Interatomic Potential”. J. Phys. Chem. B 122 .38 (Sept. 2018). pp. 8998–9006. doi: 10.1021/acs.jpcb.8b06476.

This talk is part of the Machine learning in Physics, Chemistry and Materials discussion group (MLDG) series.

This talk is included in these lists:

• Hanchen DaDaDash
• Lennard-Jones Centre external
• Machine learning in Physics, Chemistry and Materials discussion group (MLDG)
• virtual ZOOM meeting ID: 263 591 6003, https://zoom.us/j/2635916003

Note that ex-directory lists are not shown.