Skip to content

Thermal Transport across SiC–Water Interfaces

Summary

Nonequilibrium classical molecular dynamics (NEMD) was used to study thermal transport across interfaces between 3C–SiC and water, including crystallographic plane, surface termination (C vs Si), and hydrophilic versus hydrophobic solid–liquid coupling tuned via Lennard-Jones cross-terms. The work relates thermal boundary conductance (TBC) to interfacial liquid structuring and reconciles anisotropic TBC behavior across terminations using the density depletion length as a descriptor.

Methods

  • Engine and ensemble: Simulations used LAMMPS with a staged protocol: energy minimization; NVT equilibration at 300 K for 1 ns (Nosé–Hoover thermostat, 100 fs time constant); NVE for 1 ns to check stability; then imposed heat addition/removal in 1.5 nm-thick slabs for 5 ns; production sampling of kinetic energy and coordinates every 10 ps over 7.5 ns. Time step 1 fs (SHAKE on rigid water).
  • Water model: SPC/E; long-range electrostatics via PPPM (accuracy \(10^{-6}\)); rigid water enforced with SHAKE.
  • SiC: MEAM potential for the solid; bulk SiC thermal conductivity estimated by extrapolation from finite slabs (reported ~368 W/m·K, ~5% above experiment), supporting use of the parametrization for the present analysis.
  • Solid–liquid coupling: 12–6 Lennard-Jones Si–O and C–O only; ε\(_{Si–O}\) varied to span wetting while σ\(_{Si–O}\) = 2.63 Å and C–O parameters fixed (e.g. σ\(_{C–O}\) = 3.19 Å, ε\(_{C–O}\) = 0.005 eV, 13 Å cutoff) to match reference wetting behavior.
  • Geometry: Two SiC slabs with water between; (100) and (111) orientations; periodic in x,y; fixed outer layers in z; slab dimensions ~10 nm length, transverse areas ~2.62×2.62 nm\(^2\) (100) and ~2.78×2.67 nm\(^2\) (111); ~6 nm water gap; ~1000–1100 molecules to keep bulk pressure consistent across wetting states.
  • TBC: Imposed heat flux 5–15 nW; TBC \(G\) from J = G ΔT\(_\mathrm{int}\) using linear response of temperature profiles; interface ΔT from extrapolation of liquid and solid temperature profiles.

Findings

  • Interface-level TBC depends on plane, termination, and wetting, but does not universally track wettability alone; interfacial liquid structuring (layering, high-density zones) must be considered alongside bonding strength.
  • Density depletion length reconciles anisotropic TBC and termination-dependent behavior; the TBC–depletion scaling differs between Si and SiC, so a single universal TBC–depletion fit across materials is not supported.
  • For pristine planes, G ∼ 1 + cos(θ)-type scaling can hold per atomic plane, with anisotropic heat transfer between planes; icelike or ordered water regions can enhance TBC when present.
  • Mechanisms and comparisons: Interfacial heat flow is dissected relative to liquid layering and wetting strength; bulk MEAM SiC thermal conductivity is checked against experiment (~5% deviation noted in Methods) as a sanity check before interpreting interface TBC.
  • Sensitivity and outlook: TBC shifts with ε\(_{Si–O}\) wetting scans, termination, and plane; authors discuss how contact-angle-like metrics alone mis-rank interfaces. Limitations of classical MEAM + SPC/E + LJ coupling are summarized under Limitations below. Numbers and figures should be verified in the ACS Appl. Mater. Interfaces PDF (pdf_path).

Limitations

Empirical MEAM + SPC/E + LJ coupling limits transferability to reactive or charged interfaces; NEMD flux and finite-size choices affect extracted G and profiles.

Relevance to group

Classical solid–liquid thermal transport workflow comparable in spirit to interface studies in reactive systems; no ReaxFF parameterization in this paper.

Citations and evidence anchors

  • DOI: 10.1021/acsami.8b10307 (ACS Appl. Mater. Interfaces).