Silica surface states and their wetting characteristics

Summary¶

The study couples VASP PBE-PAW calculations with Grimme D3 van der Waals corrections (500 eV cutoff, 4×4×1 k-mesh for slab models) to Amber classical MD using ClayFF for silica and SPC/E water. Surfaces include talc (001), tridymite (001), quartz (001), and a siloxanated quartz model built by fitting a talc-like siloxane overlayer onto quartz (001) followed by DFT relaxation. Together the methods quantify wetting, hydroxylation propensity, interfacial layering, and cluster hydration energies that bracket non-polar siloxane versus polar silanol chemistries. The mineral set spans phyllosilicate and framework silica polymorphs so hydrophobic basal planes can be contrasted with more hydrophilic cuts without changing the water model.

Methods¶

1 — Classical MD (sessile drops, interfacial water)¶

Engine / code: Amber (including Sander), ClayFF for silica and SPC/E for water. Systems / geometry: (001) talc, tridymite, quartz, and a DFT-informed siloxanated quartz model; for contact-angle “sessile drop” runs, ~140×140 Å² surfaces, >15 Å slab thickness, 3D PBC, ~1300 water molecules, ~6 nm droplet diameter. Ensemble / thermostats: NVT at 298 K; 1 ns equilibration then 200 ps for drop contours; the paper also reports a separate 1 ns + 1 ns NVT protocol for interfacial number-density and retention analyses. Time step and thermostat: N/A — not reported in the article text (standard Amber defaults implied). Real-space / LJ cutoff: 10 Å. Barostat / servocontrol of pressure: N/A — NVT cells. Electric field: N/A. Replica / enhanced sampling: N/A — direct MD only. Output: contact angles from 2D density contours averaged over 200 ps; Wenzel-style discussion where noted.

2 — DFT (VASP)¶

Functional: PBE; PAW; 500 eV plane-wave cutoff; Grimme D3 for vdW; 4×4×1 k-mesh (for (001) slab models tabulated in the Methods). Structures and observables include relaxed (001) terminations, H₂O chemical-potential / hydroxylation-style probes (e.g. four H₂O/nm² in ~30 Å vacuum in the DFT part of the Methods), plus E_ads trends used to set Amber+ClayFF chemistry of siloxanated quartz.

3 — ReaxFF / RMD and FF training¶

N/A — the paper does not use reactive RMD or a ReaxFF reparameterization; it couples PBE+vdW VASP with ClayFF+SPC/E Amber NVT dynamics at room temperature.

Findings¶

Outcomes and mechanisms. Non-polar siloxane-terminated models are more hydrophobic (contact angles ~80°, ~3 Å depletion at the interface, weaker layering). Silanol-rich faces are more hydrophilic with brighter 1–2 Å layering and cluster hydration energies ~−1.2 to −1.6 eV, in line with immersion calorimetry the paper cites.

Comparisons to experiment / literature where stated: the energetics trends are framed against immersion data (see citations in the source), not a full world wetting database.

Sensitivity to processing history (surface preparation). Calcination vs re-hydroxylation switches terminations (siloxane vs silanol), so wetting responses and flotation-relevant behavior shift for the same nominal mineral cleavage face.

Corpus / KB honesty The Amber+rigid water treatment does not follow proton-transfer RMD; Siloxanated quartz is an intermediate morphology when industrial grains are not one ideal termination everywhere.

Limitations¶

ClayFF with rigid SPC/E water does not capture dissociative hydrolysis pathways that the authors address only in selected DFT calculations. See Findings above for corpus-level caveats vs authored claims.

Relevance to group¶

The paper is a silica–water interface benchmark for geochemical and mineral-processing classical MD, complementary to reactive ReaxFF studies elsewhere in the corpus.

Citations and evidence anchors¶

papers/Others/Jiaqi_Jin_Surface_Innovations_2019.pdf

reaxff-family

DFT + ClayFF pipeline: see reaxff-family for contrast with reactive water–oxide field development.