Skip to content

Atomic insight into tribochemical wear mechanism of silicon at the Si/SiO2 interface in aqueous environment: Molecular dynamics simulations using ReaxFF reactive force field

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 Applied Surface Science article investigates tribochemical wear of single-crystal silicon against amorphous silica in an aqueous environment using ReaxFF molecular dynamics implemented in LAMMPS, with George Psofogiannakis and Adri C. T. van Duin contributing reactive force-field expertise from Penn State alongside Tsinghua- and Southwest Jiaotong-affiliated coauthors. The motivation is practical: silicon-based microelectromechanical systems operate in humid environments, and silicon chemical mechanical polishing relies on nanoscale removal mediated by slurry particles and water chemistry. Atomic force microscopy experiments cited in the paper suggest that interfacial Si–O–Si bridge bonds play an important role in silicon removal during aqueous wear, but the atomistic sequence of bond-breaking events remains incompletely resolved. The simulations aim to identify explicit silicon removal pathways and to separate chemical oxidation effects from mechanical load effects at a sliding Si/SiO\(_2\) contact.

Methods

1 — MD application (ReaxFF / LAMMPS). Engine: LAMMPS with a ReaxFF parametrization formed by merging the Si/Ge/H and H\(_2\)O parameter sets cited in §2 (Supporting Information documents the validation path). Geometry: five-layer slab-on-slab model (4.26 nm × 4.26 nm × 8.0 nm cell) with PBC in x and y; bottom rigid Si fixture, mobile Si(100) layer, 300 H\(_2\)O molecules, mobile amorphous silica film, and top rigid silica slab that can move laterally and vertically to impose load. Surface prep: separate ReaxFF equilibrations of Si(100)+water and SiO\(_2\)+water to build H/OH/H\(_2\)O-passivated surfaces (Figures S3b / S5b), then combined into the sandwich wear cell (Figure 1). Wear protocol (five steps): (i) approach silica until the target normal force is reached, (ii) hold normal load on the top rigid slab along z, (iii) 250 ps equilibration of the compressed stack, (iv) 1 ns lateral slide along +x at 10 m/s (0.1 Å/ps), (v) 35 ps normal retraction at 10 m/s. Normal loads: uniform stresses equivalent to 2.0, 4.0, and 6.0 GPa applied to the top rigid body (§2). Ensemble: NVT with Nosé–Hoover thermostat, 300 K, 0.25 fs timestep, 100 fs damping, Verlet integration (§2 cites prior Yeon practice). Barostat: N/A — load is imposed mechanically via the rigid top slab rather than a hydrostatic barostat. Electric field: N/A — not used. Replica / enhanced sampling: N/A — not used.

2 — Force-field training. N/A — the manuscript combines published ReaxFF libraries; it does not report a new parameter fit in the main text.

3 — Static QM. N/A — not used for the production wear trajectories (DFT literature cited only for mechanistic context in Discussion).

4 — Analysis. OVITO visualization of Si removal counts, Si–O–Si bridge populations, H\(_2\)O consumption, and friction traces (Figures 3d, 8).

Findings

Removal mechanisms. The abstract and §3 describe two Si removal channels: (i) rupture of stretched Si–O–Si bridges on the Si side, assisted by H attaching to bridging O; (ii) fracture of Si–Si bonds within Si–Si–O–Si chains that are mechanically loaded through interfacial Si–O–Si bridges to silica.

Pressure response. Under the 0.3 bond-order molecular recognition criterion quoted in §3, no Si removal occurs at 2.0 GPa, whereas 7 and 14 Si atoms are removed at 4.0 and 6.0 GPa, respectively, tracking the growth of interfacial bridge-bond populations (Figure 7 narrative).

Role of water. Water both oxidizes the Si near-surface (contributing Si–O–Si in the O-depth profiles of Figure 3c) and, mechanically, can separate surfaces—yielding the dual chemical/mechanical interpretation summarized in §4 (see also Figure 8 friction / water-count correlations).

Comparisons / context. The Introduction ties the model to AFM-scale Hertzian contacts and CMP slurry silica particles; Discussion notes remaining velocity gaps between MD and AFM experiments.

Corpus honesty. Detailed snapshots and bond-order time series are in Figures 2–8 of the PDF (pdf_path).

Limitations

  • Time and length scales remain below engineering CMP loads; qualitative mechanism focus.

Relevance to group

Collaborative Tsinghua/PSU effort with van Duin on ReaxFF tribochemistry at oxide interfaces.

Citations and evidence anchors

  • DOI: 10.1016/j.apsusc.2016.08.082 (papers/Jialin_Wen_AppSurfSci_2016.pdf).
  • Text-aligned pointers: normalized/extracts/2016wen-applied-surf-atomic-insight_p1-2.txt

Reader notes (navigation)