Modeling and simulations for 2D materials: a ReaxFF perspective
Summary¶
This 2D Materials review (DOI 10.1088/2053-1583/acd7fd) surveys how ReaxFF reactive molecular dynamics has been applied across the two-dimensional materials ecosystem: layered van der Waals systems, non-layered two-dimensional phases, and heterostructures where interfaces dominate performance. The article is led by Nadire Nayir with Adri van Duin as senior author and is explicitly aimed at readers who need a methods-level map—what ReaxFF can and cannot do—rather than a single application note. The local corpus PDF is a galley filename (papers/Nayir_2D_materials_ReaxFF_review_2023_galley.pdf); cite the DOI for bibliographic identity.
Methods¶
4 — Review / perspective (2D Materials). Same DOI and bibliography as [[2023nadire-nayir-2d-materials-modeling-simulations]]; this file uses the alternate galley path Nayir_2D_materials_ReaxFF_review_2023_galley.pdf. The Methods of the work are “read the 2D ReaxFF literature and synthesize parametrization/validation, RMD sampling (including *metadynamics/parallel-tempering/force-biased* MC), and ReaxFF+continuum/phase-*field* links where cited primary groups used them. N/A to quote a universal LAMMPS .in file—every tunable ReaxFF+MD number (timestep, NVT/NPT ramps, QEq frequencies, collective-*variable* metadynamics) is in a cited source, not a one-*row* table in this review**.
Force-field (review only). N/A (new ReaxFF fit in this manuscript), but N/A (lack of discussion): the text does cover how ReaxFF is trained and tested (QM reference set, classical opt vs ML-*assisted* pipelines) in 2D-*materials* stories (see per-*section* bibliography). N/A (standalone DFT-*only* “Methods” as a separate workstream)—it is a review (plus cited DFT/DFT+*continuum* trends in cited work).
Corpus honesty. Local galley PDFs can carry line numbers; prefer a final IOP VOR file (same DOI) for citation-grade pagination and figures (see the sibling slugs under this DOI + the corpus non-primary-article catalog when hashing/deduping RAG objects).
ReaxFF fit (reviewed literature only). The article surveys how cited groups build ReaxFF parameter sets from DFT reference data (PBE-class and hybrid QM in cited work), optimize with parabolic or ML-assisted pipelines, and validate on reaction sets and experimental benchmarks in cited publications—N/A to reproduce a unique training data set in this review file.
Cited RMD (slot coverage). Representative ReaxFF studies use molecular dynamics in LAMMPS with 3D periodic boundary conditions, slab supercells with \~10⁴ atoms, 0.2–0.25 fs timestep, NVT or NPT ensembles, Berendsen or Nose–Hoover thermostat damping, 1 bar NPT where used, ps equilibration, ns production runs, and temperature setpoints from 300 K upward; N/A for a universal E-field or umbrella sampling recipe across every cited work.
Findings¶
Where ReaxFF helps¶
Most useful for reactive edge/defect/growth chemistry and environmental exposure scenarios across graphene, TMDs, MXenes, hBN, III–V 2D, heterostructures.
Caveat¶
Transferability is system-dependent—validate per chemistry.
Outlook¶
Highlights simulation–experiment gaps for in situ synthesis pathways and need for richer datasets.
Limitations¶
Breadth trades off against depth in any single material system; readers should follow primary citations for numerical benchmarks. Galley vs final layout may differ cosmetically.
Reproducibility notes¶
Readers implementing review recommendations should capture ReaxFF parameter file names and training-element coverage when reproducing cited studies—2D materials often reuse C/H/O subsets with different bond corrections than bulk oxides. For accelerated sampling, record bias parameters and collective variables if used, since rare-event statistics are method-dependent.
The galley filename in this slug is a corpus convenience; editorial corrections after acceptance may change pagination—use DOI-first citations for any locator-dependent claims (figure panels, page ranges). When importing figure numbers into retrieval indices, prefer the publisher’s XML version if available, since galley figure numbering can drift during production. For teaching use, treat each subsection as a reading list: follow the cited primary paper for any quantitative claim about transferability or error bars.
Relevance to group¶
Central 2DCC / Penn State collaboration piece with Adri van Duin as senior author.