Skip to content

Mixed-pattern cracking in silica during stress corrosion: A reactive molecular dynamics simulation

Evidence and attribution

Authority of statements

Prose below summarizes the publication identified by doi, title, and pdf_path. Crack velocities and strain values are stated in the article; do not treat this page as a substitute for the paper’s tables and figures.

Summary

ReaxFF MD is used to study stress corrosion cracking (SCC) of strained α-quartz in liquid water vs a dry reference. Over the 3 ns trajectory described in the abstract, no crack growth is seen for dry quartz at 23% strain, whereas substantial propagation appears in water at lower strains (17%, 20%, 22%). Reported crack speeds in water decrease with increasing strain in that abstract’s listing, and crack-tip strain analysis ties growth to hydrolysis of strained Si–O bonds at the tip. The work also discusses mixed cracking modes combining stress corrosion and more brittle, stress-driven failure as load increases.

Methods

Force field (ReaxFF Si/O/H)

  • ReaxFF follows the van Duin reactive force-field formulation with bond-order-dependent energetics and dynamic charge equilibration appropriate for Si/O/H chemistry (introduction/methods excerpt). Parameters for Si/O/H are taken from prior benchmarking studies on bulk silica, liquid water, and silica–water interfaces cited as validated against QM and experiment in the article.

System construction

  • α-quartz supercell: 14 × 9 × 3 rectangular unit cells built from cells of ~4.91 × 8.50 × 5.42 Å containing 12 Si and 6 O each; an ~12 Å notch seeds a crack-like defect (Methods section).
  • Relaxation: the notched sample is equilibrated in the canonical (NVT) ensemble at 300 K for 50 ps before loading (extract).

Wet vs dry setups and loading

  • Wet interface: a 68 × 37.5 × 15.3 Å cell contains 1000 H₂O molecules placed with random positions/orientations at 0.99 g cm⁻³ after energy minimization (extract; minimization paragraph continues in the PDF beyond the clipped extract).
  • Dry reference: simulations without the liquid water region are compared under 23% strain in the abstract’s dry case.
  • Straining / SCC trajectories: the abstract highlights 3000 ps SCC runs with no growth for dry quartz at 23% strain, versus substantial propagation in liquid water at 17%, 20%, and 22% strain.
  • Timestep, thermostat damping, and nonbond cutoffs for production SCC runs are not stated in the checked-in normalized/extracts/2013zhang-computationa-mixed-pattern-cracking_p1-2.txt (text ends mid minimization paragraph); consult the Comput. Mater. Sci. PDF for integration settings.

Force-field training: N/A — the work uses a validated Si/O/H ReaxFF parameterization from prior silica–water literature (as cited in the article); it does not report a de novo ReaxFF fit in this paper.

1 — MD application (atomistic dynamics). Engine / code: Reactive MD with ReaxFF; integrator package N/A — not stated in the p1–2 extract (papers/ReaxFF_others/Zhang 2014 _ ReaxFF MD of stress corrosion.pdf). System: notched α-quartz supercell (14 × 9 × 3 unit cells of ~4.91 × 8.50 × 5.42 Å each, 12 Si and 6 O per cell) plus ~12 Å notch; wet cells add 1000 H₂O in ~68 × 37.5 × 15.3 Å at 0.99 g cm⁻³ after minimization (Methods / extract). Boundaries / periodicity: 3D PBC bulk with notch and optional water region; exact lattice vectors PDF-grounded. Ensemble: NVT at 300 K for 50 ps pre-strain equilibration of the notched solid (extract). Timestep / thermostat / barostat / duration for production SCC: N/A — not in the p1–2 extract beyond the abstract’s 3 ns SCC window and strain percents—see article. Temperature: 300 K equilibration; SCC strains 17–23% as in abstract. Pressure: N/A — strain-controlled mechanics rather than quoted NPT targets here. Electric field: N/A — not used. Replica / enhanced sampling: N/A — not used. Shear / shock: N/A — not the reported protocol in the abstract-level summary.

Findings

  • Over 3 ns, no crack growth appears for dry quartz at 23% strain, whereas substantial propagation occurs in liquid water at 17%, 20%, and 22% strain.
  • In water, crack speeds are reported as 7.1 m/s, 3.1 m/s, and 0.57 m/s at 17%, 20%, and 22% strain, respectively—all faster than the bulk experimental crack velocity quoted in the article for comparison.
  • Crack-tip strain maps and species analysis tie growth to hydrolysis of strained Si–O bonds at the tip; at higher applied strain, the authors describe mixed-mode response combining stress-corrosion growth with more brittle, stress-driven fracture.

Limitations

  • Nanoscale models and short times vs macroscopic SCC; ReaxFF accuracy for silica chemistry under extreme strain must be judged against the paper’s benchmarks.

Relevance to group

Stress corrosion of silica is a recurring theme in oxide reliability; ReaxFF provides an atomistic picture of water-assisted bond rupture at crack tips.

Citations and evidence anchors

  • DOI: https://doi.org/10.1016/j.commatsci.2013.09.045 (papers/ReaxFF_others/Zhang 2014 _ ReaxFF MD of stress corrosion.pdf).