The coalescence of incipient soot clusters
Scope
Reactive MD of acetylene chemistry up to nascent soot (~3.5 nm), then isothermal coalescence of isolated incipient clusters (800–1800 K); surface-area evolution used to define a characteristic coalescence time.
Summary¶
Reactive molecular dynamics begins from a thousand acetylene molecules that collide, react, and polymerize into linear and branched hydrocarbons and polycyclic aromatics until nascent soot clusters of up to about 3.5 nm diameter form. Packing density and C/H ratio track nucleation. The study then isolates incipient clusters from the reactive bath and simulates pairwise coalescence at 800–1800 K, monitoring surface area to extract a characteristic coalescence time. Smaller clusters (up to roughly 760 atoms) coalesce rapidly (within ~0.1 ns), especially at 800–1000 K, whereas higher temperatures (1200–1600 K) yield more rigid, aromatic-rich particles that coalesce more slowly. Clusters beyond about 1300 atoms do not fully coalesce within the maximum 5 ns trajectories reported. The two-stage workflow separates gas-phase nucleation from post-nucleation agglomeration, which is useful when interpreting soot growth as a sequence of chemistry-limited and diffusion-limited steps.
Methods¶
1 — MD application (reactive two-stage workflow). The authors use reactive molecular dynamics in LAMMPS (per Carbon Methods) with a bond-order ReaxFF-class or related reactive force field for hydrocarbon/soot growth. Stage A (nucleation): a gas-phase cell of 1000 C\(_2\)H\(_2\) molecules under 3D PBC evolves to branched PAH-like and incipient soot clusters of order 3.5 nm diameter, tracking C/H ratio and packing density. Stage B (coalescence): isolated incipient clusters are annealed pairwise at 800–1800 K; NVT control (see article for thermostat and timestep in fs), with trajectories up to 5 ns for the largest clusters; surface area–time analysis gives a characteristic coalescence time. Barostat / isotropic pressure control: N/A for the excerpted isochoric coalescence runs. N/A — external electric field; N/A — umbrella or metadynamics in the reported protocol.
2 — Force-field training. N/A — application of an existing reactive field, not a new parameter fit study.
3 — Static QM. N/A as the main method.
4 — Experiments. N/A — computational only.
Findings¶
- Incipient soot formation proceeds through rich gas-phase chemistry to branched aromatics and nascent particles with quantified C/H and density evolution.
- Coalescence is fastest for smaller clusters at lower temperatures in the window studied; aromatic content at higher temperatures impedes coalescence within the simulated times.
- Very large clusters (>1300 atoms) remain non-coalesced within 5 ns under the investigated conditions.
- Surface-area tracking provides a first-pass timescale characterization for coalescence as described in the paper.
- The authors relate coalescence time trends to practical soot models that must parameterize collision efficiencies between incipient particles drawn from flame samples.
Comparisons, sensitivity, corpus. The two-stage design compares nucleation chemistry to post-nucleation agglomeration; temperature strongly controls how fast small clusters coalesce. Limitations (corpus): local extract is abstract-heavy—confirm timestep and ReaxFF designation in the PDF.
Limitations¶
Force-field fidelity limits quantitative agreement with experiment; long-time agglomeration and oxidation are outside the scope. Particle charge and ion seeding, omitted here, can alter coalescence in real soot flames.
Relevance to group¶
Corpus combustion/soot paper using reactive MD (methodological neighbor to ReaxFF combustion studies).