Skip to content

Erosion of Spacecraft Metals due to Atomic Oxygen: A Molecular Dynamics Simulation

Evidence and attribution

Authority of statements

Prose below summarizes the Journal of Spacecraft and Rockets article identified by doi, title, and pdf_path.

Summary

Low-Earth orbit (LEO) environments contain atomic oxygen (AO) that erodes spacecraft materials; flight experiments are expensive, motivating atomistic models. Prior ReaxFF studies emphasized polymers; this paper asks whether ReaxFF molecular dynamics can reproduce high-energy AO impacts on silver and gold slabs representing spacecraft metals. The authors report erosion coefficients and surface temperature evolution as functions of cumulative oxygen impacts (e.g., up to 100 oxygen atoms in the abstract-level protocol) and compare simulation outcomes to literature erosion data for LEO-relevant conditions.

The introduction motivates ~4.5 eV relative collision energies as representative of LEO AO encounters for orbiting hardware, explaining why bond-breaking reactive frameworks are needed rather than purely elastic scattering models.

Methods

1 — MD application (RMD, sequential AO impacts)

This is ReaxFF reactive MD in the same spirit as Rahnamoun & van Duin and Zeng et al., but applied to Ag(001) and Au(001) slabs. Engine / program name: N/A — Journal of Spacecraft and Rockets Sec. II specifies the ReaxFF parameter sets and integration settings but does not name a LAMMPS-style package. ReaxFF parameter sets: Lloyd et al. for Ag/Ag/AgO; Joshi et al. for Au–O (citations as in the paper’s Table 1 footnotes and Sec. II.A). System size and composition: (001) FCC Ag and Au slabs of approximate lateral extent 32×32 Å and 40 Å thickness, placed in a tall simulation cell (100 Å height, PBC in x and y). Boundaries: in-plane PBC; O is introduced along +z toward the open (001) surface. Ensemble / stages: NVT equilibration with a thermostat at 200 K on the slab; during O deposition/impact the run switches to NVE and the thermostat is removed so the collision energy is not instantaneously quenched by the bath (per Sec. II.B). Time step: 0.5 fs (Table 1). O impact energy: 4.5 eV (7.4 km/s in +z). Cadence / duration: one O every ~200 fs; 100 O impacts over 20 ps (40,000 time steps) as in Table 1 / text. Barostat / NPT or mean-stress servocontrol: N/A — NVE impact segments at fixed cell volume. Shear, shock piston, and electric field: N/A. Erosion accounting: atoms beyond ~10 Å from the pre-impact surface are treated as eroded (definition tied to Ag–Ag / Ag–O bond lengths, Sec. II.B and Rahnamoun et al.-style cutoff logic).

2 — Force-field training

N/A — uses published Lloyd and Joshi ReaxFF sets; the article rehearses the ReaxFF energy partition (Sec. II.A) for readers but does not re-fit parameters.

3 — Static QM

N/A — there is no Kohn–Sham / first-principles DFT in the manuscript; the 4.5 eV AO reference energy is taken from the LEO / beam-literature Introduction context. The authors compare their ReaxFF sputter-style yields (mass per O atom) to in-orbit and Banks et al. laboratory data (Sec. III). A subsidiary NVT(200 K)-style run with a tight slab thermostat (Fig. 7) shows no Ag erosion, consistent with Rahnamoun & van Duin-style polymer work.

Findings

After 100 oxygen impacts, the silver erosion coefficient aligns closely with literature LEO values in the authors’ comparison. Gold shows minimal erosion relative to silver, consistent with its reputation for AO resistance. The authors argue ReaxFF MD offers a comparatively low-cost way to screen AO–metal interaction scenarios for materials selection and lifetime estimation, subject to force-field limitations.

Interpretation. The work positions temperature evolution as a diagnostic: hot spots coincide with reactive removal regimes, echoing themes from earlier ReaxFF AO studies on polymers and silica cited in the introduction.

Materials selection narrative. Because Ag and Au bracket high vs low erosion under the same impact protocol, the study supports using ReaxFF MD as a screening tool for new coatings or alloys before flight qualification, provided parameter scope includes the elements of interest.

Limits. The manuscript does not replace orbital environment models that include angular flux distributions, surface roughness at micrometer scale, or synergistic UV/electron exposure; those effects require multiphysics extensions beyond atomistic MD.

Limitations

Reactive metal–oxygen models may not transfer to alloys or oxide-passivated surfaces without reparameterization; impact statistics in MD are not a full orbital environment model.

Relevance to group

Demonstrates ReaxFF application to reactive gas–metal collisions outside hydrocarbon- or silica-centric group studies.

Citations and evidence anchors