Skip to content

Molecular dynamic simulation of spontaneous combustion and pyrolysis of brown coal using ReaxFF

Evidence and attribution

Authority of statements

Prose below summarizes the publication identified by doi, title, and pdf_path. Stoichiometric labels in the abstract use φ notation—verify definitions in the paper.

Summary

ReaxFF MD simulates pyrolysis and combustion of a brown coal model at large (>1000 atoms) scale. Runs explore fuel-lean, fuel-rich, and stoichiometric oxidizer conditions at high temperature to access chemistry within affordable CPU time. The abstract notes combustion initiates by thermal fragmentation, followed by H abstraction and H₂O formation; potential energy trends differ between combustion vs pyrolysis; temperature effects dominate over density in the pyrolysis cases highlighted; formaldehyde appears as an intermediate consistent with literature.

Methods

  • Force field: ReaxFF parameterization for C/H/O/N/S/B coal chemistry (Castro-Marcano-type Illinois No. 6 training lineage as referenced).
  • Pyrolysis: 16 literature brown-coal molecules in periodic cells (~60×56×44 Å) at ρ = 0.08, 0.10, and 0.20 g/cm³; minimize at 10 K (NVE), NVT equilibration 10 ps with Δt = 0.1 fs, then heat 2000–4000 K in 500 K steps (150 ps per stage, Berendsen thermostat τ = 100 fs, Δt = 0.25 fs, 150 ps production) to follow thermal decomposition.
  • Combustion: 12 coal molecules plus 250 / 500 / 1000 O₂ (equivalence ratios φ ≈ 0.5, 1.008, and 2.0, labeled fuel-lean, stoichiometric, fuel-rich in the paper) in ~93×79×69 Å periodic cells; 10 K minimization (NVE) then NVT heating 2000–4000 K in 500 K intervals (250 ps per interval, τ = 100 fs) to access combustion chemistry at accelerated temperatures.

1 — MD application (atomistic dynamics). Engine / code: LAMMPS with ReaxFF (papers/ReaxFF_others/Sanjukta_BrownCoal_Fuel_2014.pdf). Systems: pyrolysis cells with 16 literature brown-coal molecules in ~60×56×44 Å boxes at ρ = 0.08, 0.10, 0.20 g/cm³; combustion cells as above. Boundaries: 3D PBC. Ensemble: NVE minimization at 10 K, then NVT heating/production with Berendsen thermostat (τ = 100 fs). Timestep: 0.1 fs (early equilibration) and 0.25 fs (heated stages) as stated in Methods bullets above. Duration: 10 ps equilibration (pyrolysis path), then staged heating (150 ps per 500 K step pyrolysis; 250 ps per 500 K step combustion). Barostat / pressure: N/A — NVT gas-phase cells; no hydrostatic NPT target summarized here. Temperature: 2000–4000 K staged ramps. Electric field: N/A — not used. Replica / enhanced sampling: N/A — not used.

2 — Force-field training: N/A — applies a C/H/O/N/S/B ReaxFF parameterization from the Castro-Marcano / Illinois No. 6 lineage as referenced, without reporting a new fit.

Findings

  • Combustion initiates with thermal fragmentation of the coal skeleton, followed by H abstraction and O₂ chemistry that produces large H₂O counts in the monitored species populations; gas-phase yields (CO, CO₂, H₂, H₂O, residual O₂) depend strongly on the three O₂ loadings (φ ≈ 0.5, 1.0, 2.0 with 250 / 500 / 1000 O₂ molecules in §2.2).
  • Pyrolysis vs combustion differ in potential-energy drift with temperature (exothermic oxidation vs endothermic pyrolysis trends described in the Discussion), and temperature effects outweigh density effects for the pyrolysis cases highlighted.
  • Formaldehyde appears among intermediates, which the authors compare qualitatively to literature observations on oxygenated pyrolysis/combustion products.

Limitations

  • Simplified coal chemistry and high-T acceleration; quantitative agreement with experiment is partial and qualitative.

Relevance to group

Coal combustion ReaxFF benchmark in the Fuel literature alongside other fossil pyrolysis entries.

Citations and evidence anchors

  • DOI: https://doi.org/10.1016/j.fuel.2014.07.058 (papers/ReaxFF_others/Sanjukta_BrownCoal_Fuel_2014.pdf).