Skip to content

Atomic insight into tribochemical wear mechanism of silicon at the Si/SiO2 interface in aqueous environment

Summary

ReaxFF MD in LAMMPS models tribochemical wear of single-crystal silicon against hydroxylated amorphous silica with an aqueous interlayer, focusing on Si removal pathways and how normal pressure and water jointly control oxidation versus mechanical detachment. The setup mimics AFM-like contact with a rigid counterbody and explicit water to capture stress-assisted hydrolysis at Si/SiO\(_2\) junctions relevant to chemical mechanical polishing and MEMS reliability. Weiwei Zhang and Adri C. T. van Duin coauthor the reactive modeling line.

Methods

LAMMPS runs use ReaxFF for Si/Ge/H and H₂O (parametrization citations [26,27] in the article; SI for additional validation). Separate equilibration builds H/OH/H₂O-passivated Si(100) and hydroxylated silica, then stacks them with ~300 water molecules in a sandwich geometry (SI figures S3b / S5b). The wear cell is a multilayer slab-on-slab stack with 4.26 nm × 4.26 nm in-plane periodicity, a rigid bottom Si slab, and a rigid movable top silica counterface to mimic AFM-like contact; normal stress on the top body is stepped at 2.0, 4.0, and 6.0 GPa. Production segments use NVT at 300 K, Nosé–Hoover thermostat (100 fs damping), velocity Verlet, and Δt = 0.25 fs, with OVITO postprocessing. Out-of-plane cell thickness, full equilibration versus production durations, and electrostatic or QEq update details are not recovered from the short proof excerpt used here—see the Appl. Surf. Sci. article PDF. In-plane PBC applies to the wear patch; vacuum separates the stack along the interface normal as in the cited SI figures. Wear loading replaces hydrostatic NPT control; external electric fields and bias-based enhanced sampling are not used.

Force-field training. N/A — applies published ReaxFF parametrizations.

Static QM / DFT. N/A — reactive MD study.

Findings

The abstract describes two Si removal pathways: (i) breaking stressed Si–O–Si bridges on Si with H attaching to bridging O; (ii) rupturing Si–Si bonds in stressed Si–Si–O–Si chains at the interface—both removing Si via interfacial Si–O–Si bridges. The work is framed against AFM tribology on Si/silica/water where load and water strongly affect wear but experiments lack molecular resolution. Normal load on the silica counter-slab (2.0–6.0 GPa in the protocol above) increases Si removal by promoting interfacial Si–O–Si bridge formation; water both oxidizes the Si surface and spaces surfaces to reduce intimate contact. Detailed outlook beyond the abstract belongs in the version-of-record issue; this corpus pdf_path is an uncorrected proof—confirm pagination versus Appl. Surf. Sci. 390, 216–223 (2016).

Limitations

Corpus PDF is an uncorrected proof; confirm pagination against the Applied Surface Science issue. Proof-stage journal page placeholders appear in some headers. For corpus notes on proof PDFs, see NON_PRIMARY_ARTICLE_PAPER_SLUGS.md.

Relevance to group

van Duin co-authorship; ReaxFF tribochemistry for Si/SiO₂/water stacks relevant to CMP and MEMS wear discussions.

Citations and evidence anchors

Reader notes (navigation)

For CMP-relevant wear rates, combine these atomistic stress thresholds with continuum contact mechanics; this paper supplies chemistry-aware removal pathways, not engineering wear coefficients measured at wafer scale. Normal load sweeps (2.0–6.0 GPa) map to fixed displacement control on the rigid silica plate as described in Applied Surface Science.