Development and Application of a ReaxFF Reactive Force Field for Ni-Doped MoS2
Summary¶
Doping molybdenum disulfide with nickel modulates catalytic and tribological response, but transferable classical models that capture both structural polymorphism and chemistry are scarce. Mohammadtabar et al. parameterize a ReaxFF description for Ni-doped MoS₂ using a broad density functional theory training set covering substitutional and intercalated Ni sites under multiple strain modes. The field is validated against DFT structural and energetic metrics for 2H monolayers, relative stabilities of competing sites, and vacancy-bearing configurations, then applied to reactive simulations of sputtering deposition and annealing that probe amorphous-to-crystalline evolution in doped films. The motivation ties to tribological films where nickel is introduced during synthesis or wear. The combination of strain-rich training data and processing simulations distinguishes this work from static lattice-only parameterizations.
Methods¶
DFT training corpus (C)¶
- Structures: 2H-MoS\(_2\) with Ni at Mo and S substitutional sites and at octahedral / tetrahedral intercalation sites.
- Deformation sampling: Uniaxial, biaxial, triaxial, and shear strain paths on selected configurations.
- QM details: Functional, basis, k-mesh, and convergence settings appear in J. Phys. Chem. C DOI
10.1021/acs.jpcc.3c00668.
ReaxFF optimization targets (A)¶
- Fits reproduce relaxed structures, site-energy ordering (e.g. tetrahedral vs octahedral Ni), 1H vs 1T relative energies, and selected defected configurations within tolerances quoted in the paper.
Reactive MD applications (B)¶
- Deposition / annealing: Sputtering-like impact simulations followed by thermal annealing to study amorphous→crystalline conversion in Ni-doped films.
- Numerics in the VOR/SI: Timestep, thermostat, impact energy / flux for sputter-like events, and annealing schedules for the a-C → crystal post-processing; hash-linked galley here may differ in pagination from VOR.
MD application (sputter + anneal on Ni-doped film)¶
Engine / code: LAMMPS with the new Ni–MoS\(_2\) ReaxFF. Deposition/impact and NVT-type (or as stated) anneal in 3D PBC TMD supercells; T, kinetic energy of arrivals, and duration in JPCC. N/A — not an open-circuit electrolyte or SCCM reactive flow RMD as summarized; N/A — no NPT barostat unless the VOR specifies NPT for deposition stress; N/A — no replica/metadynamics; Coulomb and QEq in the VOR/SI. N/A — no shock Hugoniot NEMD; N/A — no static external electric field in the RMD protocol summarized on this page.
Findings¶
Validation¶
The fitted ReaxFF matches the DFT validation suite for Ni-doped MoS\(_2\) within the authors’ stated error bounds. The abstract lists explicit checks: ReaxFF vs DFT relaxed structural parameters, the tetrahedral/octahedral site energy gap in doped 2H, 1H vs 1T monolayer energies, and vacancy-bearing doped 2H structures.
Processing simulations¶
Sputter + anneal trajectories capture amorphous-to-crystalline pathways accessible with the new field, supporting deposition / tribology-oriented modeling of doped TMD films.
Scope and extensions¶
Gas-phase / solid training shown in the paper does not automatically validate electrochemical or solvated environments without additional benchmarks. Porting to other dopants should repeat site preference and polymorph DFT checks before trusting extrapolated chemistry.
Limitations¶
Training chemistry focuses on Ni doping sites and strain windows listed in the paper; electrochemical or strongly oxidizing environments require separate validation.
Relevance to group¶
Group-authored ReaxFF for TMD doping with applications in tribology and catalysis modeling contexts.