Development, applications and challenges of ReaxFF reactive force field in molecular simulations
Summary¶
This review (Chinese–English) surveys ReaxFF development, parameterization philosophy, and applications across multiple domains: high-energy organics, hydrocarbons and coal chemistry, nanostructured carbons and silicon, solid–liquid and biological interfaces, catalysis on metals and oxides, and electrochemistry including fuel cells and lithium batteries. It also enumerates limitations: parameter transferability, accuracy relative to QM for selected observables, and challenges representing electronically excited states within a classical reactive framework.
The article is written as a bibliographic synthesis rather than a single benchmark study: it explains the ReaxFF energy decomposition (bond, over/undercoordination, valence, torsion, conjugation, van der Waals, Coulomb) following van Duin-style formulations and uses representative citations to illustrate where reactive MD has been impactful.
Methods¶
As a review, the manuscript does not define one new production trajectory; methods are the literature scope and comparison axes. The text walks through ReaxFF energy decomposition (bonded and nonbonded contributions and variable charge practice in the van Duin-line formulation as cited), typical QM-derived training data used in parameterization, and application clusters such as oxidation, pyrolysis, solid–liquid interfaces, mechanical loading with bond rupture, and electrochemical-adjacent problems. It points readers to primary parameterization studies, software contexts such as LAMMPS where discussed, and common validation strategies. Canonical DOI for bibliography should be taken from publisher metadata if needed; the corpus doi field may be empty because this ingest is a named offprint file (FCE-0016-15034-HY).
Findings¶
The review positions ReaxFF as a pragmatic bridge between DFT-level training data (within covered reaction and phase spaces) and large-cell classical MD timelines. It summarizes success domains (e.g., oxidation, catalysis, battery- and fuel-related interfaces in cited work) while stressing limitations: transferability outside training stoichiometries, accuracy for barriers and excited states, and the need to treat parameter files as application-specific contracts rather than universal physics. Readers should extract any numerical performance claim only from the cited primary study, not from this survey’s prose alone.
Limitations¶
Review articles age quickly; software defaults, best practices, and parameter databases evolve. Any quantitative claim about a specific material must be traced to primary sources. The corpus copy is a publisher offprint; pagination may differ from other editions.
As a bilingual survey, it can help non-English readers find entry points into ReaxFF literature, but it should not be used as a canonical description of any single parameter file’s validation status.
Relevance to group¶
High-level entry for ReaxFF scope and limitations, complementary to group-specific parameterization notes and theme hubs.
Citations and evidence anchors¶
- Corpus PDF:
papers/ReaxFF_others/ReaxFF_review_FCE-0016-15034-HY-offprint.pdf. - Excerpt alignment:
normalized/extracts/2016reaxff-venue-fce-15034-hy_p1-2.txt.
Reader notes (extended)¶
If your goal is citation for a formal bibliography, resolve the journal DOI from the publisher metadata; the corpus retains the offprint PDF chiefly as a stable file path for MANIFEST provenance.