Skip to content

Study of high density polyethylene (HDPE) pyrolysis with reactive molecular dynamics

Evidence and attribution

Slug / manifest mismatch

The wiki paper_id slug references Kowalik, but the registered PDF is Liu et al. in Polym. Degrad. Stab. 2014 (HDPE pyrolysis with ReaxFF). Front matter above matches the actual article; rename paper_id only via a governed migration.

Summary

This article applies ReaxFF molecular dynamics to high-density polyethylene (HDPE) pyrolysis using a 7216-atom model system—the authors state this as a first ReaxFF application to HDPE pyrolysis at this scale in their abstract. Simulations run in the NVT ensemble for 250 ps total sampling across 2000–3000 K conditions. Primary gas formation pathways are extracted with a post-processing tool VARMD that decodes reaction events from trajectories.

The work compares simulated product distributions and global kinetics (via an ωC31 kinetic framework described in the paper) to Py-GC/MS experiments and literature data on polyethylene thermolysis. The abstract reports that predicted ~90% mass-loss times fall near experimental ranges from the literature, and that detailed reaction networks from trajectories agree broadly with prior mechanistic literature.

Methods

1 — MD application (RMD, ReaxFF, LAMMPS)

Engine: LAMMPS reax package (Polym. Degrad. Stab. 2.1.2). FF: ReaxFF in the van Duin formulation (as cited in the Introduction; C/H parameters for alkanes in that lineage). System / composition: bulk HDPE (8 PE chains, n=150, ρ≈1.0 g/cm³), 7216 atoms in a cubic PBC box (edge ~38.2 Å after DREIDING/Amorphous Cell construction as described). Ensemble: NVT; Nose–Hoover-type thermostat (paper: Nosé–Hoover with damping 0.1 ps). Timestep: 0.25 fs. Protocol: one RMD trajectory at each temperature 2000 K, 2125 K, 2250 K, 2375 K, 2500 K, 2750 K, 3000 K for 250 ps; PBC in all directions as in the PE cell. Barostat / servocontrol of mean pressure: N/A — NVT only. Electric field: N/A. Replica / enhanced sampling: N/A—single RMD trajetories at each T. Analysis: VARMD to list elementary reactions from trajectories; kinetic treatment via ωC\(_{31}\) (paper) and comparison to Py-GC/MS + literature.

2 — Force-field training

N/A — uses established ReaxFF; DREIDING appears only for initial cell build and minimization, not RMD dynamics.

3 — Static QM / DFT

N/A — the contribution is ReaxFF RMD; no DFT reoptimization in this 2014 work.

Findings

The authors present atom-resolved scission pathways for HDPE at high temperature and argue that ReaxFF MD is a practical tool for mechanism mining despite high-T acceleration inherent to short MD windows. Overall kinetics from the ωC31 analysis are used to estimate timescales for deep conversion, matching order-of-magnitude experimental thermolysis timescales cited in the abstract.

Limitations

Short trajectories and high temperatures accelerate chemistry beyond typical lab ramp rates. ReaxFF uncertainty affects branching selectivity among radical pathways. The slug mismatch complicates corpus discovery until metadata is normalized.

VARMD-style decoding is the paper’s distinguishing feature: it converts reactive trajectories into human-auditable reaction lists, which is valuable for building mechanism diagrams for polymer pyrolysis even when absolute rates remain scale-sensitive.

Relevance to group

Polymer pyrolysis benchmark using ReaxFF + VARMD-style analysis—complements hydrocarbon combustion and pyrolysis papers in the broader ReaxFF corpus.

Citations and evidence anchors

  • DOI 10.1016/j.polymdegradstab.2014.03.022; Polym. Degrad. Stab. 104 (2014) 62–70.
  • Corpus PDF: papers/ReaxFF_others/Study of high density polyethylene (HDPE) pyrolysis with reactive molecular dynamics.pdf.
  • Excerpt alignment: normalized/extracts/2019kowalik-venue-bez-tytu_p1-2.txt.

Reader notes (extended)

If you need a bibliography export, prefer doi + Polym. Degrad. Stab. metadata above rather than the filename stem, which reflects historical ingest naming rather than authorship.