Scholarlink

<p>High-capacity and cost-effective sodium (Na) metal anode receives increasing attention for constructing high-energy-density metal batteries. However, the unstable solid electrolyte interphase (SEI) that forms on Na metal anodes drives detrimental dendrite growth and capacity fade, and its formation mechanisms remain poorly understood. Herein, an accelerated on-the-fly learning (AOFL) approach is introduced to uncover the mechanistic underpinnings of SEI formation. By combining conventional on-the-fly learning with similarity structure screening, AOFL achieves 71% faster simulations than <i>ab initio</i> molecular dynamics while maintaining comparable accuracy. The ClO<sub>4</sub><sup>−</sup> decomposition forms Na<sub>2</sub>O during the interfacial reaction simulation, while proton ion from 1,2-dimethoxyethane (DME) by reactive oxygen leads to NaOH formation, both of which are identified as critical inorganic SEI components. These insights afford theoretical guidance for elucidating SEI formation mechanisms and for the rational design of advanced electrolytes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Atomic-scale tracking of sodium metal-electrolyte reactions via adaptive machine learning force fields

  • 期刊论文

  • 发表日期：

    2025年10月28日

查看全文
  • Tianchen Zhang,
  • Jiamin Pei,
  • Zhongheng Fu,
  • Yi-Lin Niu,
  • Yu-Chen Gao,
  • Yao-Peng Chen,
  • Tian Zhang,
  • Zewei Gu,
  • Nan Yao,
  • Suyu Jin,
  • Liang Li,
  • Wenxuan Fan,
  • Xiang Chen,
  • Dawei Zhang,
  • Qiang Zhang

摘要

High-capacity and cost-effective sodium (Na) metal anode receives increasing attention for constructing high-energy-density metal batteries. However, the unstable solid electrolyte interphase (SEI) that forms on Na metal anodes drives detrimental dendrite growth and capacity fade, and its formation mechanisms remain poorly understood. Herein, an accelerated on-the-fly learning (AOFL) approach is introduced to uncover the mechanistic underpinnings of SEI formation. By combining conventional on-the-fly learning with similarity structure screening, AOFL achieves 71% faster simulations than ab initio molecular dynamics while maintaining comparable accuracy. The ClO4− decomposition forms Na2O during the interfacial reaction simulation, while proton ion from 1,2-dimethoxyethane (DME) by reactive oxygen leads to NaOH formation, both of which are identified as critical inorganic SEI components. These insights afford theoretical guidance for elucidating SEI formation mechanisms and for the rational design of advanced electrolytes.