TY - JOUR
T1 - Automated development of molten salt machine learning potentials
T2 - Application to LiCl
AU - Sivaraman, Ganesh
AU - Guo, Jicheng
AU - Ward, Logan
AU - Hoyt, Nathaniel
AU - Williamson, Mark
AU - Foster, Ian
AU - Benmore, Chris
AU - Jackson, Nicholas
N1 - Publisher Copyright:
© 2021 American Chemical Society.
PY - 2021/5/6
Y1 - 2021/5/6
N2 - The in silico modeling of molten salts is critical for emerging "carbon-free"energy applications but is inhibited by the cost of quantum mechanically treating the high polarizabilities of molten salts. Here, we integrate configurational sampling using classical force fields with active learning to automate and accelerate the generation of Gaussian approximation potentials (GAP) for molten salts. This methodology reduces the number of expensive ab initio evaluations required for training set generation to O(100), enabling the facile parametrization of a molten LiCl GAP model that exhibits a 19000-fold speedup relative to AIMD. The developed molten LiCl GAP model is applied to sample extended spatiotemporal scales, permitting new physical insights into molten LiCl's coordination structure as well as experimentally validated predictions of structures, densities, self-diffusion constants, and ionic conductivities. The developed methodology significantly lowers the barrier to the in silico understanding and design of molten salts across the periodic table.
AB - The in silico modeling of molten salts is critical for emerging "carbon-free"energy applications but is inhibited by the cost of quantum mechanically treating the high polarizabilities of molten salts. Here, we integrate configurational sampling using classical force fields with active learning to automate and accelerate the generation of Gaussian approximation potentials (GAP) for molten salts. This methodology reduces the number of expensive ab initio evaluations required for training set generation to O(100), enabling the facile parametrization of a molten LiCl GAP model that exhibits a 19000-fold speedup relative to AIMD. The developed molten LiCl GAP model is applied to sample extended spatiotemporal scales, permitting new physical insights into molten LiCl's coordination structure as well as experimentally validated predictions of structures, densities, self-diffusion constants, and ionic conductivities. The developed methodology significantly lowers the barrier to the in silico understanding and design of molten salts across the periodic table.
UR - http://www.scopus.com/inward/record.url?scp=85106143469&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85106143469&partnerID=8YFLogxK
U2 - 10.1021/acs.jpclett.1c00901
DO - 10.1021/acs.jpclett.1c00901
M3 - Article
C2 - 33908789
AN - SCOPUS:85106143469
SN - 1948-7185
VL - 12
SP - 4278
EP - 4285
JO - Journal of Physical Chemistry Letters
JF - Journal of Physical Chemistry Letters
IS - 17
ER -