Skip to content

Molecular dynamics study on temperature-dependent interfacial thermal resistance and wettability at Ga-Diamond/Cu interfaces

FieldValue
JournalCase Studies in Thermal Engineering
Published2026-01-24
AuthorsJiaqing Zhang, A. A. Mohamad, Qiuwang Wang, Wenxiao Chu
DOI10.1016/j.csite.2026.107747
Citations1

The modeled Ga-Diamond interfacial thermal resistance fell 52.4% as temperature rose from 323 K to 1023 K, reaching 6.17 × 10⁻² mm² K/W. The Ga-Cu resistance fell 56.5% over the same range and remained much lower than the Ga-Diamond value.

Temperature-dependent interfacial thermal resistance

Section titled “Temperature-dependent interfacial thermal resistance”

The transport range begins at 323 K, above the cited 303 K melting point of gallium. Both interfaces had lower ITR as temperature rose, but Ga-Diamond retained much higher resistance at both endpoints.

InterfaceITR at 323 KITR at 1023 KChange over 700 K
Ga-Diamond1.296 × 10⁻¹ mm² K/W6.17 × 10⁻² mm² K/W52.4% reduction
Ga-Cu2.458 × 10⁻³ mm² K/W1.07 × 10⁻³ mm² K/W56.5% reduction

The temperature profiles also show a larger temperature drop near Ga-Diamond than near Ga-Cu. The authors interpret the falling resistance as stronger phonon coupling at higher temperature. They also propose that some high-frequency phonons split into lower-frequency modes, which improves transport efficiency. The Ga-Cu simulation agrees with the interfacial thermal resistance measured by Zhang et al. using TDTR. All values in the table are molecular dynamics results, not direct Ga-Diamond measurements.

Molecular dynamics model and equilibration

Section titled “Molecular dynamics model and equilibration”

Each transport model had solid-liquid-solid geometry, with Ga between the two solids. Periodic boundaries applied in x and y, while fixed walls defined the heat-flow direction along z. Lattice mismatch was below 5%.

The modeled Ga layer measured 18.092 × 22.983 × 135.72 ų along x, y, and z. The Cu and diamond structures measured 18.074 × 21.688 × 101.212 ų and 17.83 × 21.296 × 99.848 ų along x, y, and z. The interfacial analysis sampled a region about 2 Å thick.

MEAM described Ga-Ga and Cu-Cu interactions. Tersoff described C-C bonds in diamond. Lennard-Jones handled cross-interface forces with a 10 Å cutoff and Lorentz-Berthelot mixing. The Ga-Diamond Lennard-Jones parameters were 0.009052 eV for ε and 4.117 Å for σ. The Ga-Cu values were 0.08574 eV for ε and 3.3605 Å for σ.

The time step was 0.5 fs, and the initial temperature was 298 K. An NPT relaxation ran for 1.5 ns before a 5 ns NVT relaxation. The NPT stage let Ga adjust volume at 303 K and removed residual stress. After 0.5 ns of NVE equilibration, a 0.1 eV/ps heat flow was applied along z for 1 ns. Researchers then collected interface temperatures during a final 5 ns NVE run.

The reported ITR is the mean of three independent runs, calculated as A·ΔT/J. Here A is interface area, ΔT is the local temperature drop, and J is the applied heat flow.

For contact-angle work, a cylindrical Ga droplet was 99.5 Å in diameter and 20 Å high. The droplet substrates were 260.3 × 260.3 Å in plane, with thicknesses of 5.1 Å for diamond and 5.9 Å for Cu. After energy minimization, the droplet underwent two 2 ns NVT relaxations. The angle came from the density profile method, using a circle fitted through both contact-line points and the apex.

PDOS was calculated from atomic velocity data through a Fourier transform of the velocity autocorrelation function. Overlap factor S quantifies the shared phonon spectral weight on the two sides.

Across 700 K, S increased from 0.0044 to 0.0074 for Ga-Diamond. For Ga-Cu, S increased from 0.12 to 0.133 over the same span. The Ga-Cu spectrum therefore had much greater overlap, which the authors identify as the main reason for its lower ITR. Most overlap occurred at low frequencies. The authors therefore assign a major transport role to low-frequency phonons.

The phonon participation rate measures spatial localization of vibrational modes. A lower PPR means greater localization and less participation in thermal transport. An extensional threshold of PPR > 0.4 identified modes that can participate in heat transport.

On the Ga side, 16.1% of atoms met this threshold at Ga-Diamond, compared with 20% at Ga-Cu. The solid-side fractions were 92.4% for diamond and 22.1% for Cu. The authors interpret the higher Ga-Cu value as more gallium modes joining thermal transport. Diamond’s high PPR did not compensate for its much weaker spectral match with Ga.

Wettability changes in opposite directions

Section titled “Wettability changes in opposite directions”
SurfaceGa contact angle at 323 KGa contact angle at 1023 KChange over 700 K
Diamond101.7°111.2°9.3% increase
Cu60.2°12.3°79.6% reduction

Ga remained nonwetting on diamond from 323 K to 1023 K, but its contact angle increased. The Ga-Cu angle fell as temperature rose, indicating stronger wetting. The authors cite prior experiments that reported 129° on diamond and 60.28° on copper as consistency checks. The paper does not state the measurement temperatures for those reference values.

Calculated (001) surface energies were 6.28 J/m² for diamond and 1.44 J/m² for Cu. Ga surface tension was calculated from the stress difference with the Kirkwood-Buff method. It fell from 625 mN/m to 350 mN/m across the 700 K span.

The authors use Young’s equation to explain the opposite angle responses. For nonwetting diamond, cos θ is negative, so lower liquid surface tension can raise the contact angle. For wetting Cu, cos θ is positive, so lower surface tension lowers the contact angle. The large surface-energy difference explains why the same liquid responds differently on the two solids.

The five simulated temperatures produced a compact linear relation between ITR and cos θ for each interface. For Ga-Diamond, the fit was cos θ = 0.022 × ITR − 0.485, with R² = 0.97. For Ga-Cu, the fit was cos θ = −3.625 × ITR + 1.352, with R² = 0.98.

ITR and cos θ have a positive relation for nonwetting Ga-Diamond, but a negative relation for wetting Ga-Cu. The authors interpret interfacial atomic attraction as a shared cause. It changes adhesion work, interfacial spacing, and liquid ordering, while also changing phonon coupling.

They propose that an ordered, solid-like Ga layer can form a phonon bridge between the solid and liquid. This mechanism offers an explanation for why wetting and thermal resistance do not move together at the two interfaces. The two separate fits describe their material pairs and should not be treated as a universal wetting law.

Engineers should not treat the 323 K resistance as temperature independent when selecting a Ga-Diamond thermal joint. The contact angle on diamond rises with temperature, even while ITR falls, so contact angle alone does not rank heat-transfer performance. Among the modeled pairs, Ga-Cu offers the lower resistance and a 12.3° contact angle at 1023 K.

The evidence covers 323 K to 1023 K, so extrapolation beyond that range is unsupported. The results come from selected MEAM, Tersoff, and Lennard-Jones potentials applied to idealized crystals. The paper does not report applied pressure, bond-line thickness, or Ga volume for the transport cells. The main text also defers droplet-size and ITR size dependence to Supporting Information.

The paper reports means from three runs, but no standard deviations or confidence intervals. The Ga-Diamond interface has no direct experimental ITR benchmark in the reported comparison. The paper does not report tests of surface roughness, coatings, or finished device stacks. Those variables need separate testing before transfer to a device. Underlying data are confidential, which limits independent checking of the reported curves.

The liquid-metal thermal interface study needs a diamond solid surface for contact with liquid Ga and reports calculations for diamond (001). 6CCVD’s diamond materials catalog offers single-crystal and polycrystalline MPCVD-grown diamond plates and wafers up to 125 mm, with thicknesses from 0.1 to 500 micrometers. The catalog also lists Cu metallization, and the custom fabrication and engineering support page can specify test coupons and contact geometry for interface tests.

View Original Abstract

Thermal interface materials consisting of low-melting-point liquid metal indicate excellent thermal conductivity and good deformation capabilities, thereby displaying important role in thermal management applications under extreme conditions. This study introduces the molecular dynamics simulation to investigate the thermal transport and wetting behavior of Ga-Diamond and Ga-Cu solid-liquid interfaces. From 323 to 1023 K, the interfacial thermal resistance of both assemblies drops sharply as rising temperature boosts phonon coupling efficiency. Notably, analysis on the phonon participation ratio shows that Ga at the Ga-Cu interface indicates a high extended phonon share of 20%, which is larger than that at the Ga-Diamond system (16.1%). The interfacial phonon match illustrates dominant effect on the thermal transport difference. Moreover, the wettability analysis reveals that the contact angle θ of gallium at the two interfaces shows opposite tendency. The contact angle increases with temperature by 9.3% on the diamond surface, whereas it reduces by 79.6% on the Cu surface. In addition, a strong linear correlation between interfacial thermal resistance interfacial thermal resistance and cos θ is fitted, showing positive relation in the Ga-Diamond system and negative relation in the Ga-Cu system.

  1. 2025 - Thermal management enhancement of electronic chips based on novel technologies [Crossref]
  2. 2024 - The effect of active and passive battery thermal management systems on energy consumption, battery degradation, and carbon emissions of an electric vehicle [Crossref]
  3. 2022 - A review on thermal management of lithium-ion batteries for electric vehicles [Crossref]
  4. 2025 - Enhanced regenerative cooling performance with conformal TPMS channels [Crossref]
  5. 2024 - Thermal interface materials: from fundamental research to applications [Crossref]
  6. 2025 - Ultra-efficient heat transport across a “2.5D” all-carbon sp2/sp3 hybrid interface
  7. 2024 - Remarkable reduction in thermal contact resistance between target and heat exchanger using TIMs in isotope neutron source accelerator [Crossref]
  8. 2023 - Thermal interface materials with high thermal conductivity and low young’s modulus using a solid−liquid metal codoping strategy [Crossref]
  9. 2023 - Roadmap towards new generation liquid metal thermal interface materials [Crossref]
  10. 2015 - The apparent thermal conductivity of liquids containing solid particles of nanometer dimensions: a critique [Crossref]