An atomistic carbide-derived carbon model generated using ReaxFF-based quenched molecular dynamics
Evidence and attribution¶
Authority of statements
Prose sections below (Summary, Methods, Findings, etc.) are curated summaries of the publication identified by doi, title, and pdf_path in the front matter above. They are not new primary claims by this wiki.
For definitive numerical values, reaction schemes, and interpretations, use the peer-reviewed article (and optional records under normalized/papers/ when present)—not this page alone.
Summary¶
The work introduces quenched MD with ReaxFF to build three-dimensional carbide-derived carbon (CDC) models capturing heterogeneous porosity and short-range graphitic order beyond idealized slit pore geometries. Radial distribution functions, pore size distributions, and adsorption metrics are compared to experiment; post-quench compression adjusts pore statistics toward targets, and ring statistics emphasize abundant non-hexagonal rings in CDC vs conventional activated carbon scenarios. Vanderbilt/Drexel/ORNL/van Duin collaboration targets supercapacitor and adsorption modeling needs.
Methods¶
1 — MD application (quenched ReaxFF MD + optional NPT compression in LAMMPS). Reactive MD (ReaxFF) is run in LAMMPS to generate 3D carbide-derived carbon (CDC) models using a quenched MD (QMD) schedule from 3500 K → 3000 K on fluid-like randomized initial carbon cells (20000 atoms per sample in the reported QMD-x series, ~6.5–7.5 nm cubic periodic boxes). Quench duration is varied (5–500 ps), giving nominal quench rates ~100–1 K ps⁻¹ as described in the Computational Details section. NVT integration uses a Nosé–Hoover thermostat (~100.5 fs damping per the article text) with Δt = 0.5 fs. For selected models, a post-quench NPT compression step is applied (3000 K, ~20 000 atm in the excerpted protocol) using the Shinoda–Rahman-style NPT integrator in LAMMPS with Δt = 0.5 fs and stated damping values for temperature/pressure control, intended to shift pore-size distributions toward experimental targets without destroying short-range graphitic motifs. GCMC nitrogen adsorption (also LAMMPS) is used to compare simulated vs experimental isotherms after structural generation. N/A — electric field: not part of the described QMD workflow.
2 — Force-field training / fitting. N/A — new QM refit in this paper: the authors apply a published ReaxFF carbon parameterization (cited combination of references in the article) suitable for amorphous/porous carbon QMD workflows.
3 — Static QM / DFT. QSDFT and related experimental PDF/STEM comparisons appear in the analysis pipeline, but DFT is not the on-the-fly structure generator for the QMD samples.
4 — Review / non-simulation framing. N/A: primary modeling paper. Proof PDF (papers/Thompson_C_2017-proof.pdf) may show placeholder metadata; confirm final DOI/issue via publisher records (sibling: [[2017matthew-w-thompson-we-report-a-atomistic-carbide-derived-2]] when populated).
Findings¶
Outcomes and mechanisms. QMD models reproduce g®, pore-size distributions, and adsorption metrics that track experimental CDC data better than simplistic slit-pore baselines. Ring statistics show abundant non-hexagonal rings, distinguishing CDC models from many activated-carbon-like constructs at comparable mean pore sizes.
Comparisons. Simulated N\(_2\) isotherms and QSDFT-reduced pore metrics are compared to experimental CDC benchmarks; STEM images provide qualitative structural anchors in the article figures.
Sensitivity and design levers. Quench rate strongly controls amorphous vs ordered character (faster quenches → more stringy/amorphous; slower quenches → more six-membered rings and larger pores as summarized in the Results). Post-quench compression shifts pore-size distributions toward sub-nanometer targets while preserving much of the local bonding learned at slower quenches.
Limitations and outlook (as authored). The manuscript stresses that simulation quench rates lack a direct one-to-one mapping to chlorine-etching laboratory timescales; compression is likewise a targeting device, not a literal synthesis replica.
Corpus / PDF honesty. This page summarizes the proof PDF in-repo; operators should align DOI/pagination with the published MDPI C article when wiring automation.
Limitations¶
- DOI not captured in local manifest metadata—resolve from the journal issue when wiring bibliographic automation.
Relevance to group¶
van Duin contributes ReaxFF expertise to nanoporous carbon structure generation for energy materials.
Citations and evidence anchors¶
- Source PDF:
papers/Thompson_C_2017-proof.pdf(proof; confirm final DOI via publisher metadata).