Researchers have utilized large-scale, quantum-accurate reactive simulations combined with active learning and deep equivariant neural networks to study the evolution of interfaces in solid-state batteries. This approach, which avoids adjustable parameters fitted to experiments, revealed a previously unreported crystalline disordered phase, Li$_2$S$_{0.72}$P$_{0.14}$Cl$_{0.14}$, within the solid-electrolyte interphase (SEI). The simulations also elucidated Li creep mechanisms along the interface, which are critical for understanding dendrite initiation. AI
IMPACT This research demonstrates advanced AI techniques for materials science discovery, potentially accelerating battery development.
RANK_REASON This is a research paper detailing a scientific discovery using computational methods. [lever_c_demoted from research: ic=1 ai=1.0]
- active learning
- arXiv
- deep equivariant neural network
- Li$_2$S$_{0.72}$P$_{0.14}$Cl$_{0.14}$
- Matteo Carli
- solid-state batteries
- {\symcell}
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
