Skip to content

Dynamic observation of dendrite growth on lithium metal anode during battery charging/discharging cycles

Summary

Reactive MD with ReaxFF is coupled to EChemDID (electrochemical dynamics with an implicit electronic degree of freedom) to model Li-metal anodes under cyclic voltage with bond-making/breaking at the SEI/electrolyte interface. The protocol aims to capture morphological evolution of dead Li and dendrites during repeated charge/discharge with interfacial chemistry—notably using realistic electrode voltage driving fields rather than ad hoc uniform fields on electrolyte alone. A case study adds HF to an ethylene carbonate (EC)-based organic electrolyte, reporting that beneficial decomposition chemistry yields a protective interphase that suppresses large volume swings and parasitic degradation. The npj article frames the problem as time-resolved interface evolution: SEI species redistribute across cycles, and dendrite morphology is tracked as an emergent consequence of coupled electrochemical driving and reactive bond events, rather than as a purely geometric Li plating instability in a nonreactive electrolyte model.

Methods

  • Reactive dynamics: ReaxFF MD for bond-order chemistry at the Li/organic electrolyte interface.
  • Electrochemistry: EChemDID to propagate electrochemical variables consistent with applied cyclic potentials (see article for implementation details, timestep, and system dimensions).
  • Additive study: HF-containing EC electrolyte vs. reference cases to probe SEI formation and dendrite suppression.
  • Analysis: Trajectory post-processing emphasizes morphological markers for dead Li and filament growth tied to local composition changes in the organic phase and at Li surfaces as cycles accumulate (see npj figures for definitions).

MD + EChemDID: Engine: LAMMPS ReaxFF; Li | organic electrolyte interfacial cells with 3D PBC; NVT with ~0.25 fs (or stated) timestep; cyclic voltage / electrode potential (EChemDID) as equivalent to time-dependent bias (battery-relevant E/potential driving); ps–ns cumulative sampling; Nose–Hoover-class thermostat at set temperature in K; barostat N/A if fixed box; hydrostatic pressure N/A unless NPT; N/A replica exchange; Supercell atom counts and box geometry in the npj Methods/figures (not restated here).

Findings

Outcomes: time-resolved dendrite / dead Li morphology under cyclic EChemDID driving; HF+EC case shows SEI chemistry that passivates vs. baseline. Comparisons: versus additive-free baseline in the paper; not direct experiment morphology fit. Sensitivity: voltage cycling and local HF-catalyzed reactions. Limitations: size/time vs real cells—authored caveats in npj. Corpus: duplicate pdf_path (see Reader notes); same text for sibling 2022lee-npj-computat-dynamic-observation-2 slug.

Simulations provide time-resolved visualization of dendrite and dead-Li growth tied to interfacial reactions. The HF additive scenario shows passivating films that limit detrimental expansion and side reactions compared to the baseline description in the paper. The npj Comput. Mater. framing emphasizes cyclic electrode potentials with EChemDID so that electrolyte chemistry and Li deposition evolve under battery-relevant driving rather than static interfacial snapshots alone; readers should map reported morphology trends back to the HF-containing EC electrolyte case highlighted in the abstract versus any additive-free baseline the authors include.

Limitations

System sizes/timescales remain below macroscopic cells; quantitative voltage windows should be validated against experiment. Duplicate PDF registration in this corpus (see Reader notes) does not change the scientific text—only filename provenance.

Relevance to group

Shows ReaxFF + EChemDID workflow for Li-metal electrodeposition—aligned with battery interface modeling in the broader corpus.

Citations and evidence anchors

Reader notes (navigation)