Skip to content

Lithium-electrolyte solvation and reaction in the electrolyte of a lithium ion battery: A ReaxFF reactive force field study

Summary

This JCP paper extends ReaxFF to organic carbonate electrolyte species (e.g., EC, EMC, VC) and LiPF6-related chemistry so that Li solvation, solvent exchange, and decomposition pathways near reducing anode conditions can be studied reactively. DFT data on Li-associated initiation reactions and solvation structures augment the training set; the parametrization introduces distinct treatments so that Li⁺ and neutral Li can reproduce similar solvation energetics while differing in chemical reactivity—neutral Li drives electrolyte decomposition, modeled with a Monte Carlo-style charge/state update compatible with ReaxFF dynamics. Decomposition barriers depend on EC coordination number around neutral Li.

Broader framing in the article highlights that SEI formation from organic carbonate reduction at graphitic anodes requires models that distinguish Li⁺ solvation from Li⁰ chemistry, because nonreactive force fields cannot follow bond-breaking sequences that set onset potentials and product distributions in electrolyte decomposition.

Methods

Training data (DFT). DFT reference data cover Li-associated initiation reactions for organic carbonate electrolytes and binding energies of Li–electrolyte solvation structures, merged into the broader ReaxFF training set for optimization.

Species scope. Parameter extension targets ethylene carbonate (EC), ethyl methyl carbonate (EMC), vinylene carbonate (VC), and LiPF₆-related chemistry (as stated in the abstract).

Li⁺ vs Li⁰ representation. A second Li parameter set describes Li⁺; ReaxFF is trained so neutral Li and Li⁺ reproduce similar solvation energies, but only neutral Li initiates reactive decomposition of solvent—mirroring the physical picture that Li⁺ is chemically inert toward bond-breaking in the same way as Li⁰ in the parametrization.

Reduction / reactivity trigger. Solvent decomposition is accessed after Li⁺ → Li⁰ reduction using a Monte Carlo–type atom modification within the ReaxFF dynamics framework (per abstract).

Reactive MD. Molecular dynamics in LAMMPS-class integration with ReaxFF on finite electrolyte cells (thousands of atoms per supercell as in the paper); NVT thermostat-controlled temperature (K-scale targets in the paper); femtosecond timestep; nanoseconds of sampling / equilibration and production; PBC for bulk-like liquid; N/A — NPT barostat if runs are constant volume; N/A for GPa pressure in typical NVT electrolyte boxes. N/A for static MV/cm electric field applied across the cell in the same sense as continuum battery models—the work uses a Monte Carlo-style Li⁺/Li⁰ state change for reduction chemistry. N/A — metadynamics.

Findings

Dual-state fidelity. The fitted potential distinguishes Li atom vs Li⁺ so anode-side reduction chemistry can be represented without collapsing both into a single nonphysical state.

Barrier–solvation coupling. Decomposition reaction barriers depend on how many EC molecules solvate the neutral Li atom—highlighting local solvation number as a control on reduction kinetics in the model.

SEI context. The work supports SEI-relevant narratives where electron leakage to electrolyte reduces Li⁺ to reactive Li⁰, triggering organic carbonate decomposition pathways that fixed-bond models cannot follow.

Compared to nonreactive FFs, the model agrees in spirit with literature that reduction chemistry requires reactive sampling. Sensitivity to EC coordination around Li⁰ controls reaction barriers and thus kinetics. Limitations include transfer to full electrolyte mixtures; caveat: ReaxFF is empirical—future benchmarks against more DFT pathways may be needed. No galley-only PDF is required to read the published article; use VOR for pagination if your tree has a proof duplicate slug for the same DOI.

Limitations

  • Complex commercial electrolyte mixtures and long-time SEI evolution require careful interpretation; parameter transfer should be checked when moving to new salts or additives. For exact barriers, Monte Carlo acceptance rules, and table values, use the JCP pdf and SI; this page is a curated summary, not a substitute for those sources.

Relevance to group

Central reference for Li-ion electrolyte reactivity with ReaxFF at PSU/INL collaboration—connects to SEI formation narratives in the battery literature.

Citations and evidence anchors

  • DOI: 10.1063/5.0003333
  • Abstract: normalized/extracts/2020hossain-j-chem-phys-lithium-electrolyte-solvation_p1-2.txt

Reader notes (navigation)