Skip to content

Development of a ReaxFF potential for Pd/O and application to palladium oxide formation

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

A new Pd/O ReaxFF description is fitted to quantum data for bulk and surface properties, then applied to high-temperature MD of oxide formation on Pd(111), Pd(110), and Pd(100). A hybrid grand canonical Monte Carlo / MD (GC-MC/MD) workflow is introduced to map a theoretical oxidation phase diagram for Pd₉₃₅ clusters from 300–1300 K and very low to 1 atm O₂ pressure, reporting faster subsurface oxygen transport on (110) versus other facets consistent with literature trends (abstract; introduction opening, extract). The motivation ties Pd oxidation to catalyst operation where oxygen uptake and oxide wetting on facets influence activity—motivating atomistic models coupled to open O₂ reservoirs (introduction themes).

Methods

Grounding: papers/Senftle_JCP_PdO_2013_proof.pdf; normalized/extracts/2013senftle-venue-paper_p1-2.txt (AIP author-query cover + abstract/introduction opening).

1 — MD application (ReaxFF MD of Pd oxidation + hybrid GC-MC/MD)

  • Engine / code: Molecular dynamics simulations of oxide formation using a developed Pd/O ReaxFF potential (abstract). Specific MD software is not stated on the indexed excerpt pages.
  • System / surfaces: Pd(111), Pd(110), and Pd(100) surfaces under high temperature and pressure conditions chosen to be comparable across facets for comparing surface → subsurface oxygen transport (abstract).
  • System size & composition (clusters): The hybrid GC-MC/MD phase-diagram study targets Pd\(_{935}\) clusters (935 atoms) in contact with an oxygen reservoir as stated in the abstract (proof PDF text).
  • Hybrid sampling: A ReaxFF-based grand canonical Monte Carlo / MD (GC-MC/MD) approach samples an open O\(_2\) reservoir while relaxing structures with MD, here applied to map oxidation of Pd\(_{935}\) clusters over 300–1300 K and \(10^{-14}\)–1 atm O\(_2\) pressures to obtain a theoretical oxidation phase diagram (abstract).
  • Boundaries / periodicity: N/A — cell/cluster boundary conditions details are not stated on the indexed excerpt pages.
  • Ensemble / timestep / duration / thermostat / barostat: N/A — not stated on the indexed excerpt pages (proof pages are mostly abstract/intro).
  • Temperature: 300–1300 K window stated for the GC-MC/MD diagram study (abstract); MD oxidation simulations are described as high-temperature without a tighter numeric range on p1–2.
  • Pressure: O\(_2\) pressures \(10^{-14}\)–1 atm in the GC-MC/MD diagram study (abstract). N/A for a single hydrostatic mechanical pressure target (not the framing on p1–2).
  • Electric field: N/A.
  • Replica / enhanced sampling: GC-MC/MD is an open-ensemble sampling approach distinct from umbrella/replica exchange (abstract).

2 — Force-field training (Pd/O ReaxFF)

  • Parent FF / elements: ReaxFF Pd/O interaction potential developed within the broader ReaxFF reactive force-field framework (abstract).
  • QM reference: Parameters are fit against an extensive quantum data set for bulk and surface properties (abstract). Specific DFT functional/basis/k-mesh tables are not stated on the indexed excerpt pages—see full article Methods in pdf_path.
  • Training set / targets: Includes bulk Pd oxides, O-covered Pd surfaces, and Pd–O clusters as summarized in the abstract’s “quantum data” statement.
  • Optimization: Parameter fitting language is used at abstract level; optimizer implementation details are not stated on p1–2.
  • Reference data / validation: Validation is claimed via agreement with experiment for Pd\(_{935}\) cluster oxidation over the stated T–P window and via consistency with prior experimental surface oxidation trends discussed in the introduction excerpt (Pd(110) > Pd(100) > Pd(111) ordering appears in introduction text).

Findings

  • Outcomes & mechanisms: Oxygen penetrates to subsurface regions faster on Pd(110) than on Pd(111) and Pd(100) under the abstract’s comparable high-T/P MD oxidation conditions.
  • Comparisons: GC-MC/MD oxidation of Pd\(_{935}\) clusters matches experiment well enough across 300–1300 K and \(10^{-14}\)–1 atm O\(_2\) to support both the Pd/O potential and the hybrid method (abstract).
  • Sensitivity / design levers: Explicit temperature and oxygen pressure ranges are the thermodynamic knobs for the cluster phase diagram study (abstract).
  • Limitations & outlook: The corpus PDF is an AIP proof with author-query boilerplate; pagination and section numbering may differ from the final issue—verify locators against the published article when available.
  • Corpus honesty: extraction_quality is partial because the extract is dominated by proof metadata; quantitative MD settings and move definitions require pdf_path beyond p1–2.

Limitations

Proof-stage PDF with author queries; normalized extraction flagged partial despite readable abstract block.

Sampling: GC-MC/MD convergence depends on attempt frequencies and move mix; check SI or methods tables for equilibration lengths before drawing quantitative phase-boundary conclusions.

Relevance to group

van Duin as senior author on ReaxFF parameterization plus hybrid sampling methodology for catalytic metals.

Citations and evidence anchors

  • J. Chem. Phys. 139(4) proof header; DOI http://dx.doi.org/10.1063/1.4815820 (extract page 2).
  • Abstract text block (extract page 1–2).