Skip to content

Bubble nucleation and Leidenfrost characteristics of nanoscale porous surfaces from molecular dynamics study: Effects of surface morphology and wettability variation

Abstract

Non-equilibrium molecular dynamics with Lennard-Jones argon–platinum interactions simulates explosive boiling and Leidenfrost film formation on flat and square nano-porous platinum substrates, varying pore width (void fraction) and hydrophilic versus hydrophobic wetting of the base and pore walls; LAMMPS integrates trajectories with a 5 fs timestep after a staged equilibration and heating protocol.

Summary

The paper studies nanoscale boiling of liquid argon on platinum substrates with square nanopores, comparing flat surfaces to porous geometries with void fractions Φ = 17 %, 30 %, and 45 % (pore widths 2.4, 3.2, and 4.0 nm at fixed pore height). Hybrid wettability cases label hydrophilic (phi) and hydrophobic (pho) combinations of the flat base (B) and structured substrate (S): BϕSϕ, BϕSpho, and BphoSϕ (plus flat Bϕ / Bpho baselines).

Methods

System geometry and interactions

  • Simulation cell: 11.7 nm × 11.7 nm × 100 nm; periodic in x and y; z uses a fixed lower boundary and a reflecting top wall; long z avoids boundary interference.
  • Fluids and solid: Liquid argon slab 6 nm thick (18,522 atoms), initialized as FCC with 5.8 Å lattice constant (~1378 kg m⁻³ at 90 K). Platinum FCC with 3.92 Å lattice constant: eight base layers, bottom layer fixed, two thermostat heating layers, remaining layers conduct heat into the fluid.
  • Potentials: Lennard-Jones (12–6) for Ar–Ar, Pt–Pt, and Ar–Pt; σ_Ar–Pt from Lorentz–Berthelot mixing; ε_Ar–Pt scaled by ξ = 0.5 (hydrophobic) or ξ = 2.0 (hydrophilic) relative to ε_Ar–Ar as defined in the paper. Cutoff ≈ 4 σ_Ar–Ar.

Integration, ensembles, and software

  • Equilibration: 1 ns NVE with Langevin thermostat on the entire system at 90 K; 1 ns with Langevin removed from liquid argon only.
  • Heating: Heating layer temperature ramped 90 K → 230 K over 0.5 ns via Langevin to limit thermal shock, then held at 230 K.
  • Production: Time step 5 fs; Velocity Verlet. Simulations run in LAMMPS; visualization with OVITO (as stated). N/A — NPT barostat; N/A — hydrostatic pressure / GPa control (non-NPT heating protocol). N/A — external electric field; N/A — replica exchange in the rare-event sense (not used in this LJ model).

Findings

  • Heat transfer and Leidenfrost: Compared to flat surfaces, nano-porous substrates improve heat transfer, delay Leidenfrost film formation, and raise argon temperature at the Leidenfrost point; higher void fraction performs better among the porous designs studied.
  • Wettability ranking: BϕSϕ gives the best heat-transfer performance among hybrid layouts; BϕSpho is the least efficient; BphoSϕ is close to BϕSϕ. (See J. Mol. Liq. version-of-record for experimental and grid-refinement comparisons; this page is LJ-model-scoped.)

Limitations

Lennard-Jones argon on platinum is a minimalist model: no polar fluids, chemical reactions, or experimental temperature/pressure mapping are included.

Relevance to group

No Penn State affiliation; included as a methods / classical MD boiling reference in the broader corpus.

Citations and evidence anchors

  • DOI: 10.1016/j.molliq.2025.126867 — Journal of Molecular Liquids.