Skip to content

Molecular Dynamics Simulations of Perfluoropolyether Lubricant Degradation in the Presence of Oxygen, Water, and Oxide Nanoparticles using a ReaxFF Reactive Force Field

Summary

Heat-assisted magnetic recording (HAMR) pushes disk temperatures high enough to threaten perfluoropolyether (PFPE) lubricants, especially in the presence of humidity and oxide contaminants at the head–media interface. Industry context matters for interpreting the 1500 K simulation temperature: the goal is not to mimic steady disk temperatures but to sample rare bond-breaking events within accessible trajectory lengths, analogous to accelerated life testing in reliability engineering. Lotfi, van Duin, and Biswas simulate nine Demnum-class D4OH strands with O\(_2\), H\(_2\)O, and SiO\(_2\), goethite FeO(OH), or Fe\(_2\)O\(_3\) nanoparticles using ReaxFF MD at 1500 K to accelerate C–O and C–F bond scission chemistry. Nanoparticles appear untreated, dry-air-treated, or wet-air-treated to mimic ambient exposure pathways relevant to Western Digital-style HDD manufacturing concerns.

Methods

1 — MD application (PFPE degradation). ReaxFF simulations of nine Demnum-class D4OH strands are run with O₂, H₂O, and oxide nanoparticles (SiO₂, goethite FeO(OH), Fe₂O₃) using ADF/LAMMPS-class ReaxFF integration as described in J. Phys. Chem. C (package name appears as ADF in the abstract/methods excerpt: “All MD simulations were performed with the ADF package”). Ensemble: NVT at 1500 K for the accelerated degradation chemistry. Timestep: 0.1 fs (selected timestep in the article). Thermostat: Berendsen with 250 fs damping constant as quoted. Nanoparticles: 16 Å diameter silica example size appears in the methods discussion; surfaces are prepared untreated, dry-air pretreated, or wet-air pretreated. Trajectory lengths: representative segments include 500 000 steps (~50 ps) and 1 000 000 steps (~100 ps) at 1500 K for specific degradation stages quoted in the paper. PBC: three-dimensional PBC for gas+NP+strand supercells. Barostat / pressure: N/A — NVT gas-phase cells without NPT barostat in the excerpted protocol. Electric fields / enhanced sampling: N/A — not used.

2 — Force-field training. N/A — applies a published C/F/O/H (PFPE + oxide) ReaxFF parametrization.

3 — Experiments / continuum. N/A — motivation references HDD tribology literature rather than new experiments here.

Findings

Outcomes / mechanisms: Water strongly accelerates D4OH degradation versus O₂-only environments at the 1500 K accelerated conditions; any nanoparticle presence accelerates chemistry versus gas-only controls. Untreated SiO₂ and goethite outperform their dry/wet-air passivated counterparts in degradation rate, whereas wet-air-treated Fe₂O₃ gives the strongest acceleration among the iron-oxide cases studied.

Comparisons: trends are discussed relative to industry HDD/HAMR humidity concerns and prior PFPE degradation literature.

Sensitivity: NP chemistry, pretreatment, and humidity are primary levers; 1500 K is a computational acceleration temperature.

Limitations: device-relevant temperatures and shear are not fully captured by the high-T gas-cell protocol.

Corpus honesty: numbers above come from papers/Lotfi_CF_OxideNP_JPC_C_2018.pdf; confirm wording against your VOR PDF.

Limitations

Single high temperature and short timescales probe accelerated chemistry; quantitative rates may not map linearly to device temperatures without extrapolation.

Relevance to group

Adri C. T. van Duin co-authors; demonstrates ReaxFF for PFPE tribochemistry with oxide contaminants.

Citations and evidence anchors