Skip to content

Influence of hydroxyls on Pd atom mobility and clustering on rutile TiO₂(110)-(1×1)

Evidence and attribution

Authority of statements

Prose below summarizes the publication identified by doi, title, and pdf_path. STM images, barriers, and cluster statistics must be read from the article.

Summary

STM, DFT, kinetic Monte Carlo, and ReaxFF-based Monte Carlo study Pd adatom diffusion and sintering on hydroxylated versus clean rutile TiO₂(110)-(1×1). The abstract reports that small (1–3 atom) Pd clusters persist near 300 K on hydroxylated surfaces whereas larger clusters form on bare TiO₂ under comparable conditions; DFT gives stronger binding near OH and higher diffusion barriers on the hydroxylated surface, with ReaxFF MC extending H chemistry during aggregation (papers/Addou_Senftle_ACS_nano_2014.pdf).

Methods

Scanning probe experiments

  • STM follows Pd vapor deposition on rutile TiO₂(110)-(1×1) prepared with vs without interfacial hydroxyls; preparation includes annealing/dosing steps described in the Experimental section (Summary).

Density functional theory (kinetic inputs)

  • DFT total-energy calculations provide Pd adatom binding sites and surface diffusion barriers on pristine versus hydroxylated TiO₂(110).
  • Barrier and binding data feed kinetic Monte Carlo rates for large-scale aggregation trends (Methods for DFT code, k-points, and XC functional).

ReaxFF Monte Carlo (explicit hydrogen chemistry)

  • NVT Monte Carlo with ReaxFF extends atomistic modeling to H spillover, Pd hydride formation, and H stripping from surface OH by Pd clusters—effects not captured by DFT+kMC alone in the authors’ workflow (Summary).

Coverage note

  • Temperature, coverage, and deposition flux details belong to the ACS Nano article; do not infer numerical barriers from this wiki page alone.

1 — MD application (atomistic dynamics). Engine / code: headline dynamics are kinetic Monte Carlo on DFT rates plus ReaxFF NVT Monte Carlo for H-containing events; standalone production MD in LAMMPS/GROMACS is N/A — not the stated headline workflow on this summary layer (see article for any auxiliary trajectories). System: Pd on TiO₂(110) slabs with tunable OH; atom counts PDF-grounded. Boundaries: slab, in-plane PBC; normal vacuum gap N/A here — confirm in PDF. Ensemble: NVT MC (ReaxFF stage); DFT segments are 0 K relaxations. Thermostat: N/A — the ReaxFF stage is Monte Carlo at prescribed temperature, not continuous MD with a Berendsen/Nosé–Hoover damping parameter on this summary layer (see ACS Nano for any auxiliary MD thermostat settings). Timestep / barostat / pressure: N/A — MC and static DFT dominate the summarized protocol (no NPT hydrostatic control described here). Temperature: ~300 K cluster comparisons (abstract). Electric field: N/A — not used. Replica / enhanced sampling: N/A — not used.

3 — Static QM / DFT-only. DFT supplies binding sites and diffusion barriers for Pd on pristine vs hydroxylated TiO₂(110); functional, basis, and k-mesh choices are in the Experimental / computational sections of the article (N/A — not duplicated on this page).

Findings

  • On hydroxylated surfaces, DFT reports stronger Pd binding near OH and higher diffusion barriers, so kMC/ReaxFF reproduce small (1–3 atom) clusters that remain stable near 300 K, whereas bare TiO₂(110) yields much larger clusters under similar deposition.
  • ReaxFF MC shows sub-nm Pd clusters can adsorb H stripped from TiO₂–OH, depleting OH and removing the diffusion impediment, explaining the bimodal size distributions (small clusters at low coverage, large clusters once OH is titrated off).
  • Compared to experiment: STM cluster sizes on hydroxylated vs clean surfaces anchor the DFT/kMC/ReaxFF story (Summary).
  • Sensitivity: OH coverage and Pd flux control whether OH-stabilized small clusters persist versus coarsening (Findings).
  • Limitations / outlook: UHV models omit ambient water and support complexity (## Limitations).
  • Corpus note: numerical barriers belong to the ACS Nano PDF, not this short wiki layer.

Limitations

  • Model UHV-like surfaces vs real wet catalysts; ReaxFF accuracy for Pd–oxide is method-dependent (validated in-article to the extent described).

Relevance to group

Adri C. T. van Duin and Thomas P. Senftle link ReaxFF to supported catalyst sintering on TiO₂, a recurring oxide–metal theme.

Citations and evidence anchors

  • DOI: https://doi.org/10.1021/nn501817w (papers/Addou_Senftle_ACS_nano_2014.pdf).