Skip to content

Study of ice cluster impacts on amorphous silica using the ReaxFF reactive force field molecular dynamics simulation method

Summary

ReaxFF reactive MD simulates hypervelocity impacts of amorphous and crystalline ice clusters (150 and 128 water molecules, respectively) onto fully oxidized and suboxide amorphous silica surfaces at 1, 4, and 7 km s\(^{-1}\). The study connects sticking versus bounce, dissociation, secondary impacts on ice-covered silica, and temperature diagnostics to judge when electronic excitation might invalidate classical chemistry. J. Appl. Phys. 119, 095901 frames the problem as icy-body impingement on silica spacecraft materials—distinct from ambient water adsorption studies—where hypervelocity clusters impart shock heating and scattering not captured by static surface adsorption models.

Methods

This DOI duplicates the article curated on 2016rahnamoun-venue-study-ice; the protocol is the same ReaxFF study in J. Appl. Phys. 119, 095901. Amorphous (150 H\(_2\)O) and crystalline (128 H\(_2\)O) ice clusters strike fully oxidized and suboxide amorphous silica built from compressed SiO\(_2\) / SiO\(_{1.5}\) precursors in periodic cells (~2600–2610-atom slabs). Clusters start ~20 Å from the surface; the combined system is NVT-equilibrated at 150 K, then given normal velocities of 1, 4, or 7 km s\(^{-1}\). The text specifies NVT for this equilibration but does not name a thermostat coupling scheme in the PDF body checked here (N/A — thermostat model beyond the NVT label). Collisions are integrated in NVE for 30 ps with a 0.1 fs timestep (with a 0.05 fs cross-check for the fastest amorphous-on-suboxide case). Each condition is run three times. Second impacts quantify mass gain/loss on ice-pre-covered silica. N/A — barostat or target hydrostatic pressure — NVE impacts omit pressure control. N/A — applied electric field; umbrella / metadynamics / replica exchange — not reported.

Force-field training. N/A — applies an existing ReaxFF description.

Static QM / AIMD. N/A — not the primary modality; the text discusses classical chemistry limits versus possible electron excitation near ~10 km s\(^{-1}\).

Findings

At 1 km s\(^{-1}\), clusters deposit on the surface; at 4 and 7 km s\(^{-1}\), partial rebound appears. Limited water dissociation occurs at the higher speeds. Second impacts on iced surfaces show velocity-dependent accumulation or stripping—for example, full accumulation at 1 km s\(^{-1}\) for both impact rounds, partial stripping after a second 7 km s\(^{-1}\) hit, and mixed behavior at 4 km s\(^{-1}\) comparing crystalline versus amorphous ice. Temperature analysis argues electronic excitation is unlikely below ~10 km s\(^{-1}\) in this model, but near ~10 km s\(^{-1}\) peak temperatures can reach very high local values where ReaxFF’s ground-state chemistry assumption may break down. Amorphous vs crystalline ice clusters differ in rigidity and energy partitioning, which feeds through to secondary-impact mass balance on pre-iced silica.

Limitations

Classical reactive MD omits explicit electronic dynamics; the paper flags ~10 km s\(^{-1}\) as a prudence threshold for excitation effects.

Dust charging, plasma sheath acceleration, and multi-body impacts in orbital debris environments exceed the single-cluster normal incidence protocol summarized on this page.

Crystalline ice polymorphs and porous silica substrates may alter energy partitioning compared to dense amorphous silica films used as baselines in the article.

Relevance to group

van Duin-group application of ReaxFF to ice–silica hypervelocity collisions with explicit velocity staging.

Citations and evidence anchors