Skip to content

Erratum: Parametrization of a reactive many-body potential for Mo–S systems [Phys. Rev. B 79, 245110 (2009)]

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 numerical values, reaction schemes, and interpretations, use the peer-reviewed article (and optional records under normalized/papers/ when present)—not this page alone.

Summary

This erratum corrects implementation issues discovered when coding the authors’ Mo–S reactive many-body potential into LAMMPS. It supplies revised short-range pairwise parameters and Lennard-Jones parameters for Mo–Mo, S–S, and Mo–S, replaces the LJ tail with a cubic spline beyond 0.95σ to avoid an excessively repulsive wall, and fixes a misprinted elastic constant \(c_{12}\) for bcc Mo in the original paper (correct value 189 GPa). Additional many-body tables omitted originally are provided so third-party codes can reproduce the model. MoS\(_2\) \(c_{11}\) and bulk modulus values shift upward modestly after the correction (reported as 261 GPa and 81 GPa vs 242 GPa and 75 GPa respectively in the erratum text).

Methods

This erratum supplies corrected short-range pairwise parameters and Lennard–Jones (LJ) parameters for Mo–Mo, S–S, and Mo–S interactions for implementation in LAMMPS, and corrects a misprinted elastic constant \(c_{12}\) for bcc Mo in the parent paper (correct value 189 GPa, consistent with the bulk modulus relation \(B=(c_{11}+2c_{12})/3\)). To avoid an overly repulsive LJ wall, the LJ tail is replaced by a cubic spline beginning at 0.95σ (with σ the LJ diameter parameter) to smoothly terminate van der Waals interactions by the inner cutoff radii \(R_\mathrm{min}\). The erratum also publishes many-body tables omitted originally (coordination and sixth-order angular terms) plus an added γ[cos θ] term for S angles, needed for faithful reproduction of the potential. The corrections were developed cooperatively with University of Arkansas (Stewart, Spearot) LAMMPS implementation testing.

Force-field training (Mo–S reactive many-body; erratum to PRB 2009)

Parent FF / elements: Corrected short-range pairwise terms and Lennard–Jones parameters for Mo–Mo, S–S, and Mo–S within the authors’ Mo–S reactive many-body framework as originally published (Phys. Rev. B 79, 245110 (2009)); this erratum supplies tables needed for faithful LAMMPS implementation.

QM reference: N/A — new DFT benchmarks are not the focus of the erratum; the parent paper’s QM training context applies (see parent DOI in Citations).

Training set / reference data: N/A — additional training structures beyond the parent work are not introduced here; the erratum instead fixes implementation quantities and publishes omitted tables.

Optimization: N/A — no new global optimization campaign is reported in the erratum text (see parent publication for the original fit).

Reference data used: Elastic diagnostics at 300 K (MoS\(_2\) \(c_{33}\) at 52 GPa) motivate adjusting S–S LJ ε; revised \(c_{11}\)/bulk modulus for MoS\(_2\) are quoted after the LJ/spline fixes (261 GPa / 81 GPa vs 242 GPa / 75 GPa).

MD application (implementation note; no new trajectories in the erratum)

Engine / code: LAMMPS is the stated integration target for the corrected Mo–S potential forms.

System size & composition: N/A — no new MD supercell sizes or stoichiometries are reported in the erratum itself; users choose composition when applying the corrected tables to Mo, S, and MoS\(_2\) models.

Boundaries / periodicity: Periodic bulk and lamellar MoS\(_2\) contexts motivate the long-range vdW LJ/spline fix; explicit user PBC choices are N/A — outside the erratum’s scope.

Ensemble / thermostat / timestep / duration: N/A — the erratum does not specify NVT/NPT production settings, thermostats, timesteps, or ps/ns trajectory lengths (those belong to downstream application studies using the corrected potential).

Barostat / pressure control: N/A — not specified in the erratum beyond elastic property benchmarking language.

Temperature: 300 K appears when discussing reproducing MoS\(_2\) \(c_{33}\) after adjusting LJ parameters.

Pressure / stress: Elastic constants (GPa scale) are used to discuss mechanical property shifts after the correction.

Electric field: N/A — not applicable.

Replica / enhanced sampling: N/A — not applicable.

Findings

Outcomes / mechanisms: The authors state that revised \(R_\mathrm{min}\), \(R_\mathrm{max}\), and LJ parameters do not materially change strong short-range energetics or most mechanical predictions for Mo, S, and MoS\(_2\) relative to the original publication, while improving long-range vdW behavior in lamellar MoS\(_2\) via the cubic spline tail starting at 0.95σ.

Comparisons: After correction, predicted \(c_{11}\) and bulk modulus for MoS\(_2\) increase to 261 GPa and 81 GPa, respectively, from 242 GPa and 75 GPa in the original table (as quoted in the erratum).

Sensitivity / design levers: The erratum notes that short-range cutoffs and LJ parameters may need adjustment depending on the system class being simulated (it cites analogous cutoff practices for REBO carbon nanotube fracture studies).

Limitations / outlook (authored tone): The piece is explicitly an erratum correcting implementation issues discovered during LAMMPS coding; users must merge these tables to avoid spurious long-range forces and misprinted elastic data.

Corpus / KB honesty: Grounded in pdf_path and normalized/extracts/2012liang-venue-paper_p1-2.txt (short excerpt); full parent-paper context lives at Phys. Rev. B 79, 245110.

Limitations

  • Users of the 2009 PRB paper must merge this erratum or risk incorrect long-range forces and elastic constants.

Relevance to group

MoS\(_2\) classical reactive potential corrections relevant alongside TMD and nanocarbon mechanics papers in the corpus.

Citations and evidence anchors

  • Transition metal dichalcogenide (TMD) classical potentials
  • reaxff-family (contrast: Mo–S many-body potential line)