Skip to content

An interatomic potential for saturated hydrocarbons based on the modified embedded-atom method

Evidence and attribution

Authority of statements

Summaries track the PCCP article (doi). MEAM is distinct from ReaxFF; the paper compares to ReaxFF and REBO for hydrocarbon benchmarks.

Summary

Develops a modified embedded-atom method (MEAM) potential for saturated hydrocarbons (alkanes), motivated by applications at hydrocarbon–metal and polymer–metal interfaces where many-body metallic models must meet reactive organic energetics without mixing incompatible formalisms. The parameterization is fit to a mixed experimental and first-principles database spanning geometries, atomization energies, diatomics and molecular dimers, and high-pressure PVT for methane. The authors benchmark MEAM against second-generation REBO and ReaxFF on shared targets to situate accuracy for saturated species, explicitly noting ReaxFF as a comparator for reactive hydrocarbon modeling in the corpus context.

Methods

Evidence in this section is grounded in papers/Others/MEAM_hydrocarbons_PCCP_2014.pdf and the indexed abstract/introduction in normalized/extracts/2014meam-venue-rsc-cp_p1-2.txt (Phys. Chem. Chem. Phys. 2014, 16, 6233–6249; DOI 10.1039/C4CP00027G).

1 — MD application (atomistic dynamics)

The indexed excerpt centers on MEAM development and benchmarks versus REBO/ReaxFF; it does not spell out a single production molecular dynamics protocol (engine, timestep, NVT/NPT lengths) in the pages captured here.

  • Engine / code: N/A — not stated in 2014meam-venue-rsc-cp_p1-2.txt (consult the full PCCP article for any explicit MD package names).
  • System size & composition: Training references span alkane homologs through n-octane isomers, diatomics/dimers, and a dense methane PVT point (density 0.5534 g cm⁻³ in the abstract text)—not a single fixed supercell count in the excerpt.
  • Boundaries / periodicity: N/A — not stated in the indexed excerpt (likely PBC for bulk PVT targets in the paper body).
  • Ensemble / timestep / duration / thermostat / barostat: N/A — not stated in the indexed excerpt for benchmark MD details.
  • Temperature: 0 K atomization energies appear in the training list; high-pressure PVT data for methane are also cited as training targets (abstract).
  • Pressure: PVT pressure–volume–temperature data for dense methane are part of the reference set (abstract).
  • Electric field: N/A — not stated.
  • Replica / enhanced sampling: N/A — not stated in the indexed excerpt.

2 — Force-field training

  • Parent FF / elements: Modified embedded-atom method (MEAM) for saturated hydrocarbons, positioned as an extension of the MEAM formalism traditionally used for metals and metal compounds (introduction in extract).
  • QM reference: The abstract states MEAM is informed by density functional theory alongside pair potentials in the parameterization philosophy; functional, basis, and k-mesh specifics for each training datum are N/A — not in the short extract—read PCCP Methods.
  • Training set: Fit targets listed in the abstract/extract include (1) n-alkane bond distances/angles and 0 K atomization energies through n-octane isomers, (2) H₂, CH, C₂ diatomic curves, (3) dimer curves for (H₂)₂, (CH₄)₂, (C₂H₆)₂, (C₃H₈)₂, and (4) PVT for dense methane at the quoted density.
  • Optimization: Parameterization by fit to the mixed experimental and first-principles database (abstract); optimization algorithm/software details are N/A — not in indexed excerpt.
  • External reference data: Comparisons to second-generation REBO and ReaxFF on shared training observables; experimental PVT for methane, ethane, propane, and butane series are discussed as validation targets in the abstract narrative.

3 — Static QM / DFT-only

N/A as standalone block beyond the DFT-informed MEAM fit statement above; detailed QM settings for each training item live in the PCCP article body/SI.

Findings

Outcomes and mechanisms

The abstract states MEAM reproduces the listed experimental and/or first-principles data with accuracy comparable to or better than REBO or ReaxFF on those sets, predicts PVT for methane–butane-class systems reasonably, and gives reasonable energetics for C–C and C–H bond-breaking in saturated species.

Comparisons

Head-to-head benchmarks against second-generation REBO and ReaxFF are a core comparison axis in the abstract framing.

Sensitivity and scope levers

The abstract explicitly warns the parameterization does not accurately treat unsaturated hydrocarbons and should not be applied there—an authored scope boundary for olefinic/aromatic chemistry.

Limitations and corpus honesty

For interface or polymer–metal use cases, the introduction motivates MEAM as a bridge potential, but transferability must respect the saturated hydrocarbon training scope. Quantitative barriers, elastic data, and any MD validation tables should be taken from the full PDF rather than this wiki summary alone.

Limitations

Does not accurately treat unsaturated hydrocarbons; MEAM scope and transferability follow the paper’s caveats. For interface or shock studies that require bond rearrangement beyond saturated alkane chemistry, readers should compare against ReaxFF or other reactive models trained for those reaction classes rather than extrapolating this MEAM set. Transfer to metal interfaces likewise requires explicit metal parameter blocks compatible with the hydrocarbon subset published here. Long alkane melting and interface friction studies should verify cutoffs and neighbor lists against the PCCP implementation notes bundled with the potential release.

Relevance to group

Benchmark context for reactive hydrocarbon modeling where corpus work often uses ReaxFF; useful cross-method reference.

Citations and evidence anchors

  • https://doi.org/10.1039/C4CP00027G — Abstract and PCCP volume/page as in front matter.