Characterisation of pyrolysis kinetics and detailed gas species formations of engineering polymers via reactive molecular dynamics (ReaxFF)
LAMMPS ReaxFF pyrolysis of HDPE, PMMA, and HIPS with elevated-temperature accelerated kinetics is compared to thermogravimetric measurements and used to extract volatile and char compositions relevant to fire and combustion modeling.
Summary¶
The article develops a reactive MD workflow to follow thermal decomposition of three engineering polymers from solid-like slabs to volatile products and char. ReaxFF (reax/c in LAMMPS) captures bond breaking and reformation during pyrolysis. Simulated mass-loss/char trends and species distributions are compared to TGA experiments across heating rates, and kinetics parameters are extracted from MD and compared to experimental DTG-based analysis.
Methods¶
- Software: LAMMPS with the reax/c package for ReaxFF potentials; initial amorphous packings prepared with PACKMOL, pre-annealed using PCFF with Nosé–Hoover thermostat/barostat in stated preparation stages.
- Initial structures: Compressed to 3 nm × 3 nm × 4 nm cells (final densities \(\approx\) 0.93, 1.18, and 1.08 g cm\(^{-3}\) for HDPE, PMMA, HIPS); periodic boundary conditions in x and y with reflecting (non-periodic) z for the free surface; 12 nm vacuum along \(z\) (final box 3×3×16 nm).
- Pyrolysis protocol: NVT simulations at fixed temperatures 1300–4000 K for 300 ps at each target \(T\); 1300 K additionally run for 2 ns to capture slow chemistry. Nosé–Hoover thermostat relaxation time 0.5 ps. Bottom 0.2 nm of the slab fixed (excluded from thermostat); reactive region 3.8 nm thick. Timestep: N/A — the article text in the indexed PDF does not quote a numerical integration timestep; confirm in the full PDF or SI if needed.
- Experiment: TGA/DTA under N\(_2\) from room temperature to 800 °C at 5–30 K min\(^{-1}\) on commercial sheet samples; kinetics extracted via Kissinger–Akahira–Sunose and iterative DTG fitting (per Methods).
- Barostat / pressure control: N/A — NVT pyrolysis with no NPT production stage in the protocol summary above. Electric field and umbrella / metadynamics / replica exchange: N/A.
Findings¶
- MD reproduces major TGA/char trends for HDPE, PMMA, and HIPS and yields C\(_1\)–C\(_3\) alkane-rich volatile pools as principal fuel gases from all three polymers under the modeled conditions.
- Char composition and relative char yields from MD align with TGA-derived char mass fractions within the stated comparisons (see manuscript tables).
- Pyrolysis activation energies and pre-exponential factors extracted from MD decomposition curves are reported alongside experimental kinetics (Table 3 in the article).
The article positions reactive MD as a complement to macroscopic TGA: it resolves bond-breaking sequences and volatile speciation that are not uniquely determined from mass-loss curves alone, while still requiring elevated temperatures in simulation to match laboratory time scales.
Limitations¶
Elevated temperatures accelerate chemistry relative to experiment; absolute timescales and quantitative toxic minor species should be cross-checked for regulatory use. ReaxFF parameter accuracy varies by bond class.
Relevance to group¶
Exemplar ReaxFF + LAMMPS polymer pyrolysis study with experimental validation—useful benchmark for decomposition and gas-release workflows in the corpus.