Skip to content

Enabling Computational Design of ZIFs Using ReaxFF

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

This J. Phys. Chem. B article lays out a ReaxFF-first modeling strategy aimed at computationally assisted design of zeolitic imidazolate frameworks (ZIFs), emphasizing bond-making/breaking events that matter during processing, mechanical loading, or guest–framework interactions that fixed-bond MOF-FFs may mishandle. Coauthors span Penn State reactive simulation (Yang, Shin, van Duin) and glass / disordered framework expertise (Mauro, Bennett), reflecting the paper’s materials-informatics angle on ZIF chemistry. The ingested file is an ACS proof PDF for the JPCB article. ZIF processing can involve solvent evaporation, mechanical compaction, or defect annealing steps where lattice FFs that freeze topology miss failure modes that ReaxFF can represent.

Methods

ReaxFF molecular dynamics (ZIF chemistry)

  • Force field / code: ReaxFF parametrization for Zn–Co–C–H–O–N ZIF chemistry taken from ref 38 in the article, implemented in LAMMPS (normalized/extracts/2018yang-j-phys-chem-enabling-computational-2_p1-2.txt confirms LAMMPS usage and 0.25 fs integration).
  • Initial structures: crystalline ZIF-4, ZIF-62, and ZIF-77 configurations from the Cambridge Structural Database (CSD); equilibrated at 10 K with a Berendsen thermostat before heating sequences.
  • Thermal protocol (overview): ZIF-4 and ZIF-62 are heated to 300 K in 2.5 ps under NPT with Nosé–Hoover thermostat/barostat, then ramped toward melting (1500 K within 12.5 ps in the excerpted Methods paragraph). ZIF-77 is heated from 10 K to 900 K at the same heating-rate convention stated in the article for that composition.

  • Glass formation (ZIF-4): after high-temperature sampling, the ZIF-4 melt is quenched to 300 K within 12.5 ps in NPT to form an a_gZIF-4 glass (timings from the Methods excerpt).

  • Boundaries / periodicity: bulk ZIF crystals and melts use three-dimensional periodic boundary conditions (PBC) in LAMMPS on CSD-sourced cells, matching the VOR protocol on [[2018yang-j-phys-chem-enabling-computational]].

Validation and analysis

  • Comparisons target experimental glass metrics and selected first-principles MD benchmarks where cited (see article for tables/figures); bond-order and coordination diagnostics are used to monitor framework disordering, ligand decomposition, and melting events.

  • Replica / metadynamics / applied E-field: N/A — not used for the ZIF melt–quench ReaxFF workflows summarized here.

Findings

  • ZIF-4 melt–quench: ReaxFF reproduces key structural signatures of the glassy product (a_gZIF-4), including density, thermal response, and pore morphology, in strong agreement with experiment and FPMD references summarized in the abstract.
  • ZIF-62 melting: simulations capture how electronic vs steric effects of the benzimidazolate substituent shift melting temperature relative to ZIF-4-like chemistry (abstract-level claim; see article for quantitative T_m discussion).
  • ZIF-77 (nitro-functionalized linker): ReaxFF extrapolations suggest electron-withdrawing −NO₂ motifs can lower Zn–N-linked melting trends, but the framework is prone to oxidation/decomposition, making it a poor glass former in practice—highlighting a stability vs processability trade-off for computational screening.
  • Take-home for design loops: because ReaxFF captures bond breaking/forming during high-temperature and disordered ZIF states, it supports screening scenarios (glass formation, defect engineering, chemical resilience) that are inaccessible to nonreactive MOF force fields—at the cost of parameterization specificity to the trained chemistries.

Limitations

  • MOF/ZIF parameterization is chemically specific; transfer to new linkers/metals requires re-fitting or validation.
  • Proof PDF pagination may differ slightly from the final issue page numbers.

Curation note: this slug registers ACS proof bytes per the non-primary catalog; for stable bibliography and figure numbering, prefer the VOR PDF from DOI 10.1021/acs.jpcb.8b08094 when available locally. ZIF processing simulations in the article should be read alongside MOF-FF elastic baselines when scoping non-reactive versus reactive budgets. Penn State reactive simulation authorship (Yang, Shin, van Duin) signals continuity with broader MOF failure studies in the corpus.

Relevance to group

Foundational ZIF + Reaxff methodology paper in the group’s porous framework portfolio.

Citations and evidence anchors

Reader notes (navigation)