Machine Learning Enhanced Electrochemical Simulations for Dendrites Nucleation in Li Metal Battery

Taiping Hu,Haichao Huang,Guobing Zhou,Xinyan Wang,Jiaxin Zhu,Zheng Cheng,Fangjia Fu,Xiaoxu Wang,Fuzhi Dai,Kuang Yu,Shenzhen Xu

2024-11-06

Abstract:Uncontrollable dendrites growth during electrochemical cycles leads to low Coulombic efficiency and critical safety issues in Li metal batteries. Hence, a comprehensive understanding of the dendrite formation mechanism is essential for further enhancing the performance of Li metal batteries. Machine learning accelerated molecular dynamics (MD) simulations can provide atomic-scale resolution for various key processes at an ab-initio level accuracy. However, traditional MD simulation tools hardly capture Li electrochemical depositions, due to lack of an electrochemical constant potential (ConstP) condition. In this work, we propose a ConstP approach that combines a machine learning force field with the charge equilibration method to reveal the dynamic process of dendrites nucleation at Li metal anode surfaces. Our simulations show that inhomogeneous Li depositions, following Li aggregations in amorphous inorganic components of solid electrolyte interphases, can initiate dendrites nucleation, accompanied by dead Li cluster formation. Our study provides microscopic insights for Li dendrites formations in Li metal anodes. More importantly, we present an efficient and accurate simulation method for modeling realistic ConstP conditions, which holds considerable potential for broader applications in modeling complex electrochemical interfaces.

Materials Science

What problem does this paper attempt to address?

The paper attempts to address the issue of low Coulombic efficiency and severe safety hazards caused by uncontrollable lithium dendrite growth in lithium metal batteries. The formation mechanism of lithium dendrites is crucial for improving the performance of lithium metal batteries. Traditional molecular dynamics (MD) simulation tools are difficult to capture the lithium electrochemical deposition process due to the lack of constant potential (ConstP) conditions. Therefore, this paper proposes a ConstP method that combines machine learning force fields and charge equilibration methods to reveal the dynamic process of dendrite nucleation on the surface of lithium metal anodes.

Specifically, the study addresses this issue through the following steps:
1. **Propose a new simulation method**: Combining machine learning force fields (MLFF) and charge equilibration methods (QEq), a highly efficient constant potential (ConstP) molecular dynamics simulation scheme was developed.
2. **Simulate the electrochemical interface**: This method was applied to simulate the electrochemical interface between ethylene carbonate + lithium hexafluorophosphate (EC+LiPF6) electrolyte and lithium metal electrode, directly observing and studying the phenomena of dendrite nucleation and dead lithium aggregation during the charging/discharging process.
3. **Analyze the dendrite nucleation mechanism**: By quantitatively analyzing MD trajectories, it was revealed that the local aggregation of lithium atoms in the amorphous inorganic solid electrolyte interface (SEI) layer is the key mechanism that triggers uneven lithium deposition and dead lithium aggregation.

In summary, this study not only provides microscopic insights into lithium dendrite formation but also proposes an efficient and accurate simulation method suitable for modeling complex electrochemical interfaces.