Skip to content

Low-temperature carbonization of polyacrylonitrile/graphene carbon fibers: a combined ReaxFF molecular dynamics and experimental study

Summary

ReaxFF MD of oxidized polyacrylonitrile (PAN) with graphene explores catalytic edge and N/O chemistry during accelerated-temperature trajectories; experiments compare PAN vs PAN/graphene carbon fibers carbonized at 1250 °C vs 1500 °C. The simulation narrative emphasizes graphene edges and heteroatom groups as nucleation seeds that accelerate alignment of carbon rings and graphitic growth sequences. Experiments report large tensile strength and Young’s modulus gains for graphene-containing fibers carbonized at 1250 °C relative to neat PAN fibers carbonized at 1500 °C, framing energy/cost benefits of lower-temperature processing enabled by graphene. The process message is not merely “add graphene,” but use graphene to catalyze graphitization chemistry so lower furnace temperatures achieve superior mechanical properties than higher-T neat PAN routes.

Methods

  • ReaxFF MD: Oxidized PAN models with graphene inclusions (see article for system construction); NVT equilibration 60 ps at 300 K; five snapshots from the last 5 ps (1 ps spacing) are heated to 2200, 2500, or 2800 K at 10 K/ps, then held 1 ns at each target temperature to sample high-temperature chemistry on MD timescales.
  • Integrator: 0.25 fs timestep (validated vs 0.10 fs at 2800 K in SI); Berendsen thermostat with 100 fs damping (as stated).
  • Experiments: Carbonization of PAN and PAN/graphene fibers with mechanical testing (values in abstract).

MD application (ReaxFF). The Carbon paper uses ReaxFF-based reactive MD in 3D PBC supercells of oxidized PAN with graphene inclusions (full stoichiometry in pdf_path). The protocol summarized here: NVT equilibration 60 ps at 300 K; five snapshots from the last 5 ps; heating 10 K/ps to 2200, 2500, or 2800 K; 1 ns holds at each target temperature; 0.25 fs timesteps (cross-checked vs 0.10 fs at 2800 K in SI); Berendsen thermostat, 100 fs damping. Barostat: N/A — NVT stages as written. Pressure: N/A — not a barostat study in the MD protocol quoted. Electric field: N/A — not used. Umbrella / metadynamics / replica exchange: N/A — not used.

Findings

  • MD mechanism narrative: graphene edges plus N/O functionality act as catalytic seeds promoting ring alignment and graphitic cluster formation relative to oxidized PAN alone.
  • Experiments: PAN/graphene fibers at 1250 °C show ~91% higher strength (632 → 1207 MPa) and ~101% higher modulus (88 → 177 GPa) compared to PAN-only at 1500 °C (abstract values).
  • The 2200–2800 K MD windows are acceleration strategies; the paper uses them for qualitative reaction pathways, not as direct furnace timelines.

Corpus / versions: the uncorrected-proof PDF is tracked as [[2020rajabpour-venue-paper]]; this page uses the VOR pdf_path above for table-ready mechanics numbers when they diverge.

Limitations

Simulations use elevated temperatures to accelerate chemistry vs furnace protocols; quantitative one-to-one mapping to factory carbonization schedules requires caution. Fiber mechanics also depend on misalignment, voids, and surface defects not represented in atomistic reaction pathways; treat experimental modulus and strength as structure-sensitive metrics even when chemistry trends align. Graphene loading fractions and dispersion quality in precursor fibers can shift effective catalysis relative to ideal MD geometries.

Relevance to group

Penn State chemical/mechanical engineering co-authorship with van Duin; UVA mechanical testing collaboration.

Reader notes (navigation)