To do that, we need to understand the underlying mechanism.​
We want a descriptor to identify electrides from database.
Electrides are materials with (electron-deficient, distance-optimized)multicentered bonding formed by orbitals of surrounding atoms*.
*Simlar idea were also proposed by Xiao Dong and Artem R. Oganov, (2017), Electrides and Their High-Pressure Chemistry, In G. G. N. Angilella, Antonino La Magna (editors) Correlations in Condensed Matter under Extreme Conditions, Springer Press, and by M. Hanfland et al. ,New high-pressure phases of lithium, Nature, (2000), 174-178, 408.
Theory
σ
\sigma
σ2∗​
\sigma_2^*
σ1∗​
\sigma_1^*
s
s
tensile strain
Resonance-like
Metallic-like
Descriptor-ELF
Electron localization function (ELF): Probe the localization information of a quantum state at a point in space using Pauli exclusion principle.
ELF is a scalar field just like charge density
ELF = 0 → delocalized; ELF = 1 →localized
Ref: Andreas Savin, et al., ELF: The Electron Localization Function, Angew. Chem. Int. Ed. Engl., 1997,36,1808-1832
Ref: A. D. Becke and K.E. Edgecombe, A simple measure of electron localization in atomic and molecular systems, J. Chem. Phys., 1990, 92, 5397.
To identify the multicentered bonding:
Charge density maxima @ interstitial site
ELF maxima @ interstitial site
ELF-isosurface
Known 2D electride: Yâ‚‚C
ELF gives the location of the multicentered orbital.
But how do translate this interpertation into an automated process?
