Skip to content

Direct observation of realistic-temperature fuel combustion mechanisms in atomistic simulations

Summary

Reactive molecular dynamics with ReaxFF can follow bond rearrangements in fuel chemistry, but nanosecond horizons have historically forced >2000 K simulations for appreciable alkane reactivity. This Chem. Sci. article applies CVHD + ReaxFF in LAMMPS to extend effective timescales: n-dodecane pyrolysis is reported down to 1000 K (long ~57 ms effective trajectory in their Table 1 example), while lean oxidation reaches 700 K with an effective time up to ~39 s. CVHD grows a history-dependent bias on bond-distortion collective variables (hyperdynamics-style acceleration with metadynamics-like hill deposition). The authors argue product compositions and dominant pathways are strongly temperature-dependent and consistent with experiments and kinetic models for the comparisons shown—see [[2016bal-2016-chemsci-doi-venue-direct-observation]] for the fully tabulated MD/CVHD protocol grounded on this same DOI.

Methods

This ingested PDF (Bal_Neyts_CVHD.pdf) is a DOI-identical sibling of [[2016bal-2016-chemsci-doi-venue-direct-observation]] (different bytes on disk). The scientific protocol is the same Chem. Sci. 2016, 7, 5280–5286 article: LAMMPS + colvars + CVHD, ReaxFF (Chenoweth et al. with QEq), Δt = 0.1 fs, Langevin thermostat thermalization then Nosé–Hoover-chain NVT (0.1 ps relaxation) and Martyna–Tobias–Klein barostat NPT (1 ps relaxation) where used. System composition & PBC: pyrolysis = 912 atoms (24 n-dodecane) in a 50 × 50 × 50 ų periodic cell; lean combustion = 390 atoms (5 n-dodecane + 100 O₂) in a 40 × 40 × 40 ų periodic cell; CVHD uses the C–C/C–H strain collective variables and Gaussian hill schedules quoted on the sibling page, plus additional NPT CVHD at 1000 K (10–500 bar) with a 0.5 ps hill stride. Electric field: N/A — not used. Replica / umbrella sampling: N/A — acceleration is CVHD only.

For a reader-facing, line-by-line protocol narrative (including CVHD numerical knobs), use [[2016bal-2016-chemsci-doi-venue-direct-observation]]—this duplicate-ingest page stays short by design.

Force-field training

N/A — application of literature ReaxFF, not a new parameterization (same as sibling).

Static QM / DFT

N/A — same as sibling (no central DFT trajectory study for the CVHD runs).

Findings

The duplicate PDF does not change the scientific conclusions summarized on [[2016bal-2016-chemsci-doi-venue-direct-observation]]: CVHD unlocks low-T chemistry on second-scale effective times for oxidation (700 K) and millisecond-scale windows for pyrolysis (1000 K), with strong temperature dependence of products and mechanism (high-T β-scission / C₂-rich vs lower-T isomerization / heavier alkenes in pyrolysis; O₂-abstraction-driven vs pyrolysis-first limiting cases in oxidation). Use the sibling page + pdf_path for figure-level detail.

Limitations

CVHD introduces approximations distinct from brute-force MD; validation requires scrutiny of collective variables and time reconstruction. ReaxFF accuracy limits quantitative barrier heights and branching ratios for some channels.

Relevance to group

Methodological reference for time-scale acceleration with reactive force fields in hydrocarbon oxidation networks.

Citations and evidence anchors