Study of thermal conductivity of ice clusters after impact deposition on silica surfaces using the ReaxFF reactive force field
Evidence and attribution¶
Authority of statements
Summaries track Phys. Chem. Chem. Phys. DOI 10.1039/C5CP05741H and the local PDF/normalized/extracts/2015rahnamoun-physical-che-study-thermal_p1-2.txt abstract and introduction.
Summary¶
Ice accretion on aircraft and spacecraft surfaces motivates quantifying heat transfer in impacted ice layers: thermal transport depends on ice microstructure and how clusters attach to substrates. The paper investigates thermal conductivity (TC) of amorphous and crystalline ice after high-velocity deposition on silica using ReaxFF molecular dynamics. It states that a dual-thermostat (non-equilibrium) method computes TC, with validation by comparing crystal and amorphous ice TC against experimental references before studying deposited clusters. The abstract reports that TC drops for both ice phases after deposition on silica, more strongly on suboxide silica surfaces; crystal ice retains higher TC than amorphous after accumulation, yet 1 km s⁻¹ impacts disrupt crystallinity so TC approaches amorphous-like values; ionic species embedded in ice further reduce TC in the modeled scenarios.
Methods¶
Reactive molecular dynamics with ReaxFF uses the silica–water/ice interaction parametrization referenced in the article (bond-order + charge equilibration suitable for proton transfer at the interface). System / composition: amorphous ice clusters of ~500 H₂O molecules and crystalline clusters of 512 H₂O are prepared, equilibrated at 150 K, placed ~20 Å from amorphous silica slabs with contrasting oxidation (fully oxidized vs suboxide), then impacted at 0, 0.5, or 1 km s⁻¹. Boundaries: periodic boundary conditions in supercells constructed for thermal conductivity extraction after attachment. Ensemble / thermostat: equilibrium MD (EMD) segments and non-equilibrium MD (NEMD) with dual thermostats (direct / heating–cooling-rate style method described in the article) are used to obtain TC values. Timestep, production durations, bath temperatures, and supercell dimensions appear in Computational Methods and are not in the short local extract.
Barostat / hydrostatic pressure: N/A — the summarized workflow emphasizes NEMD heat flux rather than NPT control.
Electric field / metadynamics: N/A.
MD engine: N/A — the indexed extract refers to the ReaxFF reactive molecular dynamics program generically; the Methods section of the PDF names the implementation.
Findings¶
After deposition on silica, calculated thermal conductivity drops for both crystal and amorphous ice relative to free-standing references; the drop is more pronounced on suboxide silica. Crystal ice remains more conductive than amorphous after accumulation unless 1 km s⁻¹ impacts disorder the crystal toward amorphous-like TC. Ionic dopants inside ice further reduce TC in the abstract’s tests. Comparisons: the abstract states validation of the NEMD protocol against experimental TC for crystal and amorphous ice before deposition studies. Sensitivity: TC depends on impact speed, substrate oxidation, and doping. Limitations: ReaxFF phonon physics and NEMD finite-size effects are expected caveats (see article discussion).
Limitations¶
ReaxFF captures reactive chemistry but phonon physics may differ from spectral or GK-based estimates; NEMD introduces thermostat and size effects. Impact simulations explore a limited speed set; experimental ice microstructures may span beyond the modeled clusters.
Relevance to group¶
van Duin-co-authored ReaxFF application connecting ice–silica attachment and suboxide chemistry to thermal conductivity metrics relevant to aerospace icing and surface energy modeling—useful alongside other water/silica transport notes in the corpus.
Citations and evidence anchors¶
DOI 10.1039/C5CP05741H — PCCP paper; papers/Rahnamoun_PCCP_Ice_clusters_2015.pdf; extract normalized/extracts/2015rahnamoun-physical-che-study-thermal_p1-2.txt.