Understanding chemical bonding in adatom configurations on the surface of the magnetic Weyl semi-metal Co<sub>3</sub>Sn<sub>2</sub>S<sub>2</sub> with machine learning, 3D scanning tunneling spectroscopy, and first-principles calculations
ORAL
Abstract
The cleaved surface of the magnetic Weyl semi-metal Co3Sn2S2 offers an ideal testing ground for understanding the chemical bonding of various adatom configurations that remain after cleavage. Changes in local electronic tunneling spectra can reveal different chemical bonding states of the adatoms, however deciphering these sometimes, subtle, changes are not straightforward. In fact, the elemental composition of the adatoms themselves has recently been under debate by several groups1,2. Here we develop a machine learning workflow that combines supervised learning in the spatial domain and unsupervised learning in the energy domain that, with direct comparison to first-principles calculations performed in-tandem, allows an enhanced understanding of the elemental composition and chemical bonding nature of the adatoms and their configurations. Our findings indicate that adatoms belonging to the same structural configuration differ in electronic structure, implying a difference in chemical bonding and possibly elemental composition, paving the way towards a more complete understanding the surface of this unique material.
1. Yin, J.-X. et al. Nature Physics 15, 443–448 (2019).
2. Morali, N. et al. Science 365, 1286–1291 (2019).
1. Yin, J.-X. et al. Nature Physics 15, 443–448 (2019).
2. Morali, N. et al. Science 365, 1286–1291 (2019).
–
Presenters
-
Kevin Roccapriore
- Oak Ridge National Lab