Skip to content

Simulation Protocol for Prediction of a Solid-Electrolyte Interphase on the Silicon-based Anodes of a Lithium-Ion Battery: ReaxFF Reactive Force Field

Evidence and attribution

Authority of statements

Prose sections below (Summary, Methods, Findings, etc.) are curated summaries of the publication identified by doi, title, and pdf_path in the front matter above. They are not new primary claims by this wiki.

For definitive reaction pathways, stoichiometries, and numerical values, use the peer-reviewed Letter (and Supporting Information where cited)—not this page alone.

Summary

This Letter develops a ReaxFF reactive model aimed at SEI-relevant chemistry for liquid carbonate electrolytes in contact with Si-based anodes (pristine Si and SiO\(_x\)), including multi-solvent mixtures and VC/FEC additives, trained from ab initio reference data. Large-scale reactive molecular dynamics then samples electrolyte decomposition and interfacial reactions so that gas-phase species (for example C\(_2\)H\(_4\), CO, CO\(_2\), CH\(_4\), C\(_2\)H\(_6\)), inorganic fragments (Li\(_2\)CO\(_3\), Li\(_2\)O, LiF), and organic motifs (ROLi and ROCO\(_2\)Li with R = methyl or ethyl) can be discussed in the same framework as experimentally relevant liquid formulations—at computational cost tractable for multicomponent cells.

Methods

A — Force-field training / fitting: ReaxFF adjusted to DFT targets for carbonate reduction, ring-opening, EC one-electron routes (SI expands tables); Li-containing electrolyte chemistry for Si/SiO\(_x\) interfaces.

B — Molecular dynamics / sampling: LAMMPS via iBat battery workflow; velocity Verlet, 0.5 fs, NVT 800 K, thermostat Nose–Hoover (0.01 fs⁻¹ damping per manuscript; the Letter prints Nosé–Hoover). Cells: pristine Si vs SiO\(_x\), mixed carbonates, VC/FEC additives. Barostat / pressure: N/A — the summarized production trajectories are NVT liquid/interface cells at 800 K without NPT hydrostatic pressure control; confirm any NPT or stress-controlled segments in SI if present.

C — DFT / static QM: DFT reference data for parameter fitting and spot checks (SI).

D — Review / non-simulation framing: Primary JPCL Letter—not a review.

System / boundaries: Electrolyte–electrode supercell compositions, 3D PBC choices, and fixed vs mobile Si layers are tabulated in the Letter/SI rather than duplicated here. Duration / staging: The Letter reports multi-ns cumulative NVT reactive sampling at 800 K to accumulate SEI-relevant decomposition chemistry; exact per-segment ps/ns splits and any equilibration windows should be read from pdf_path/SI (indexed abstract alone is insufficient for full protocol tables).

Findings

On Si(100) with dangling Si sites, EC can undergo ring-opening chemistry even without Li in the shown interfacial setup, evolving C\(_2\)H\(_4\) and CO among other products, whereas on a hydrogen-terminated Si surface no EC dissociation is observed in the paired comparison—consistent with the experimental trend cited. Across anode chemistries, electrolyte composition (including mixtures) and additives shift relative product distributions, supporting the authors’ argument that a mixture-capable reactive protocol is needed for SEI modeling beyond single-solvent, fixed-composition cells.

Comparisons / limitations. The Letter explicitly contrasts dangling vs H-terminated Si reactivity and ties additive / solvent composition trends to experimental SEI-relevant observations; 800 K acceleration is a documented limitation relative to room-temperature battery operation.

Corpus / PDF honesty. Use pdf_path and SI for stoichiometric product tables beyond this summary.

Limitations

  • Elevated 800 K MD is used to accelerate reactions; it is not a direct room-temperature operating snapshot, and barrier crossing statistics may differ from lower-temperature electrochemical conditions.
  • ReaxFF remains an approximate reactive model; quantitative agreement with DFT or experiment should be checked case by case, especially for subtle additive effects and long-time SEI maturation.
  • Electrochemical reduction in the full sense (potentials, explicit electron transfer at controlled bias) is not reproduced atomistically in the Letter’s MD framing; the work focuses on chemical reactivity pathways in condensed-phase setups aligned with prior reduction-focused literature.

Relevance to group

Independent Korea Institute of Science and Technology / Sejong University work on ReaxFF + LIB SEI chemistry on Si anodes; useful cross-reference for reactive electrolyte–electrode interface modeling alongside group battery and ReaxFF lineage pages.

Citations and evidence anchors

  • DOI: 10.1021/acs.jpclett.7b00898
  • Text pointers: normalized/extracts/2017kang-seop-yun-j-phys-chem-simulation-protocol_p1-2.txt; Supporting Information for extended DFT and parameterization tables as referenced in the Letter.

Reader notes (navigation)