Unveiling Carbon Ring Structure Formation Mechanisms in Polyacrylonitrile-Derived Carbon Fibers
Scope
Joint experiments and multiscale modeling (ReaxFF atomistics plus microscale simulation) on polyacrylonitrile-derived carbon fibers, linking carbonization temperature to ring-structure evolution and mechanical trends.
Summary¶
Carbon fibers from polyacrylonitrile precursors dominate commercial production; their properties depend sensitively on stabilization and carbonization schedules. The study isolates carbonization temperature while holding other processing variables fixed, combining detailed materials characterization with atomistic ReaxFF modeling and microscale simulation to relate six-membered ring formation and graphitic ordering to strength, ductility, and modulus trends measured experimentally. The multiscale framing links atomistic ring-condensation statistics to continuum mechanical property targets relevant to aerospace and structural composites supply chains.
Methods¶
Experiments. Oxidized PAN precursor fibers (ρ ≈ 1.401 g cm⁻³, commercial source per article) were carbonized at 1800 K, 2300 K, or 2800 K for 30 min with heating rate 10 K min⁻¹ in flowing N₂ (38 scfh at the facility noted in the paper). Cooled fibers yielded measured densities 1.746, 1.772, and 1.858 g cm⁻³ at the three temperatures, respectively.
Characterization. FE-SEM and HR-TEM for morphology; CHNS/O elemental analysis; XRD (Cu Kα, 2θ = 10°–80°); Raman (514 nm); XPS (surface chemistry); gas pycnometry for density (ASTM D382-referenced protocol in the article).
Atomistic simulation (ReaxFF). Reactive simulations probe carbon ring formation and graphitic ordering kinetics as a function of carbonization temperature, tied to the experimental temperature ladder.
Microscale simulation. Continuum/microstructure-level models (per article) connect evolving ring statistics and morphology to strength, ductility, and Young’s modulus trends.
Integration. The workflow tests whether accelerated graphitic nucleation and growth at higher T explains lower strength/ductility and higher modulus measured on fibers.
ReaxFF protocol (MD application). The article reports reactive MD in LAMMPS using a CHON-capable ReaxFF description to capture carbonization-driven ring formation; N/A — full timestep, thermostat, ensemble (NVT vs NPT), and cell boundary conditions for each annealing stage are not copied line-by-line to this page (see ACS Appl. Mater. Interfaces text and any SI). N/A — barostat for hydrostatic pressure if only NVT stages are used. N/A — external electric field. N/A — replica exchange, metadynamics, or umbrella sampling for the ReaxFF block.
Static QM (block 3). N/A — not a DFT-only paper; DFT is not the central method block for the main claims.
FF training (block 2). N/A — applies an existing ReaxFF parameterization; no new ReaxFF fit is the focus of the reported work.
Findings¶
Experiments vs simulation. Mechanical testing matches the simulation directionality: strength and ductility decrease while Young’s modulus increases as carbonization temperature rises.
Mechanism. Atomistic and microscale results attribute the trade-off to faster graphitic phase nucleation and growth at elevated temperature, linking six-membered ring development and ordering to the measured stiffening and embrittlement trends.
Outlook. The authors frame the coupled experiment + ReaxFF + microscale pipeline as a basis for alternative precursors and processing optimization toward lower-cost, high-performance carbon fibers. Six-membered ring population is emphasized as a process fingerprint: higher T accelerates graphitic ordering, raising stiffness at the expense of ductility in the combined modeling story.
Limitations¶
Fiber experiments integrate many microstructural degrees of freedom; atomistic models idealize chemistry and cannot capture full furnace-scale transport or defect distributions. Microscale fracture models depend on morphology meshes derived from imaging; misalignment between 2D sectioning statistics and 3D void networks can shift predicted modulus trends even when atomistic chemistry is qualitatively correct. Raman I(D)/I(G) trends should be interpreted alongside laser spot averaging and fiber texture as described in the experimental sections.
Relevance to group¶
Flagship collaborative effort between Virginia and Penn State groups using ReaxFF alongside microscale fracture simulations for carbon-fiber process science.
Citations and evidence anchors¶
- https://doi.org/10.1021/acsami.9b15833