Skip to content

Effects of water on tribochemical wear of silicon oxide interface: Molecular dynamics study with reactive force field (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 Langmuir article reports molecular dynamics simulations with the ReaxFF reactive force field to elucidate atomistic mechanisms of tribochemical reactions at a sliding interface between fully hydroxylated amorphous silica and oxidized silicon, with the amount of interfacial water treated as a controlled variable. The motivation is nanoscale friction and wear of silicon materials in humid environments, including atomic force microscopy observations in which silicon wear can depend strongly on relative humidity and in which Si–O–Si bridge formation across contacting surfaces has been proposed as a key tribochemical channel. The simulations adopt a slab-on-slab contact intended to mimic experimentally relevant contact widths while allowing explicit control of water confined under periodic boundary conditions. Jejoon Yeon, Adri C. T. van Duin, and Seong H. Kim connect these reactive trajectories to the broader literature on humidity-dependent silicon tribology.

Methods

1 — MD application (ReaxFF tribochemistry). Force field: Fogarty et al. Si/O/H\(_2\)O ReaxFF set (cited as the authors’ Si/O/water choice). Engine / code: LAMMPS reactive MD. Geometry: slab-on-slab contact between hydroxylated amorphous SiO\(_2\) and oxidized Si(100) in a 3.19 × 3.19 × 7.0 nm\(^3\) periodic cell (~3.2 nm in-plane width to mimic AFM-scale contacts). Surface preparation: a-SiO\(_2\) prepared by melt/quench (3000 K → 300 K); Si(100) oxidized with 300 O\(_2\)-equivalent oxygen insertions at 300 K for 400 ps (100 fs damping noted for that oxidation segment); both slabs hydroxylated with 300 H\(_2\)O at 300 K NVT until silanol densities plateau (SI Figures S1–S2, Table S1). Shear protocol: (i) vertical compression, (ii) equilibration, (iii) 1 ns lateral slide at 10 m/s (0.01 nm/ps along x on the top rigid silica block), (iv) 100 ps pull-off at 20 m/s along z. Normal load: 1 GPa maintained on the top rigid body during sliding (mimicking cited AFM studies). Water coverage: 0, 20, 50, or 100 interfacial H\(_2\)O molecules (100 ≈ full monolayer, 50 ≈ 60–70%, 20 ≈ 30–40% of the interface in their cell). Temperatures: 300, 500, and 700 K cases reported for thermal sensitivity. Ensemble: NVT with Nosé–Hoover thermostat for all MD segments described. Timestep: 0.25 fs (explicit statement in Simulation Methods). Barostat: N/A — load is imposed mechanically (1 GPa target) rather than a stress-fluctuation barostat. Electric field: N/A — not used. Replica / enhanced sampling: N/A — not used.

2 — Force-field training. N/A — the study uses the published Fogarty parametrization without refitting here.

3 — Static QM. N/A — not used.

4 — Experiments cited. N/A as new experiment — AFM/tribology literature motivates boundary conditions only.

Findings

Dry interface. With no interfacial water, ReaxFF-MD shows substantial cross-interface atom transfer during 1 GPa sliding; mixing begins with dehydroxylation followed by Si–O–Si bridges spanning the a-SiO\(_2\) and oxidized Si(100) slabs (abstract).

Submonolayer water. 20 or 50 H\(_2\)O molecules enable dissociative pathways that still produce Si–O–Si bridges and sustain transfer, consistent with humidity-assisted tribochemistry discussed for AFM wear.

Near-monolayer water. 100 H\(_2\)O (~full monolayer in their cell) suppresses transfer because Si sites at the sliding plane are hydroxyl-terminated instead of forming as many interfacial Si–O–Si bridges (abstract).

Temperature. Additional runs at 500 K and 700 K (with the same mechanical protocol) probe how thermal activation alters reaction pathways (Simulation Methods paragraph).

Experimental context. The Introduction links these coverage-resolved mechanisms to non-monotonic relative-humidity wear trends in AFM (including high-RH separation of asperities by multilayer water).

Limitations

  • Idealized planar contact vs asperity-dominated real interfaces.

Relevance to group

Foundational Langmuir tribochemistry paper from van Duin’s group on silicon oxide sliding.

Citations and evidence anchors

  • DOI: 10.1021/acs.langmuir.5b04062 (papers/Yeon_Langmuir_2016.pdf).
  • Text-aligned pointers: normalized/extracts/2016yeon-venue-la5b04062_p1-2.txt

Reader notes (navigation)