AIREBO-M: A reactive model for hydrocarbons at extreme pressures
Summary¶
The AIREBO potential extends the reactive hydrocarbon framework with Lennard-Jones nonbonded terms, but those LJ interactions become unrealistically stiff at small intermolecular separations, producing excess repulsion under high pressure compared with density functional theory benchmarks for polyethylene and graphite. O’Connor, Andzelm, and Robbins introduce AIREBO-M, which replaces the carbon–carbon Lennard-Jones tail with a Morse form fit to graphite interlayer compression up to about 14 GPa and to post-Hartree–Fock steric reference data for small alkanes, while preserving ambient-density behavior where the original parametrization is adequate. The motivation section ties the failure mode to high-pressure polymer and graphite simulations where interchain repulsion dominates stress.
Methods¶
Force-field training (classical reactive hydrocarbon model)¶
Parent model: AIREBO retains its reactive bond-order intramolecular description; AIREBO-M replaces the intermolecular carbon–carbon Lennard-Jones tail with a Morse form so repulsion does not diverge as steeply at small separations. QM / experiment references: the C–C Morse branch is fit to x-ray measurements of graphite interlayer compression up to ~14 GPa and to post-Hartree–Fock reference data for steric interactions between small alkanes (for C–H and H–H channels where separate x-ray constraints are lacking). Optimization goal: soften the repulsive wall with an extra parameter while preserving ambient thermodynamics of the original AIREBO as far as possible (see JCP discussion).
MD application (validation simulations)¶
Classical molecular dynamics validates AIREBO-M on graphite bilayer compression and on quasistatic plus shock compression of orthorhombic polyethylene supercells (Sec. II–III, J. Chem. Phys.). Systems use 3D periodic cells with atom counts, lattice settings, and uniaxial shock or quasistatic loading paths spelled out in those sections; timestep, thermostat use away from the shock piston, and segment durations are given numerically there rather than duplicated on this wiki page. Barostat / NPT: N/A for the shock Hugoniot-style validation, where volume follows the shock driver instead of a hydrostatic barostat. Replica / umbrella sampling / applied electric fields: N/A for the benchmarks summarized in the abstract. Thermal / temperature control: quasistatic and shock protocols span ambient reference states and shock-heated high-pressure regimes along the Hugoniot-style paths tabulated in the article, so there is no single fixed “production temperature” to quote outside those multi-state tables.
Findings¶
Versus parent AIREBO: replacing the intermolecular C–C LJ tail removes the spurious stiffening that breaks high-pressure graphite and polyethylene benchmarks. Graphite bilayer compression from AIREBO-M matches quantum reference calculations in the tests reported in the abstract. Polyethylene: AIREBO-M reproduces quasistatic and shock compression of orthorhombic polyethylene to at least ~40 GPa in the validation highlighted in the abstract—substantially extending the pressure window where the LJ-limited form fails. Outlook (as framed in the paper): the modification is positioned as a way to keep ambient AIREBO behavior while enabling high-pressure hydrocarbon simulations relevant to shock and dense molecular solid regimes.
Limitations¶
Electronic polarization and explicit quantum accuracy are not recovered; chemistry remains within the AIREBO bond-order class.
Relevance to group¶
Comparative context for bond-order hydrocarbon models discussed alongside ReaxFF in high-pressure organics literature.
Citations and evidence anchors¶
DOI 10.1063/1.4905549.