Heat dissipation in nanoelectronics hinges on phonon behavior in thin films. Aluminum nitride is attractive for its high thermal conductivity, but at the nanoscale, scattering at boundaries and defects changes the picture. Nearby in the archive are works on anomalous thermal conductivity in nanostructures, chip cooling with films, and superconductivity in thin films, all dealing with heat or charge transport in layers. This paper stands out because it is the first to experimentally capture the transition from ballistic to diffusive regime in AlN on sapphire.
AlN films were grown by metal-modulated epitaxy on sapphire substrates, with thicknesses from 1.6 to 2440 nm. Microstructure and strain were studied by transmission electron microscopy and X-ray diffraction. Thermal resistance was measured by time-domain thermoreflectance (TDTR), a pump-probe technique with spot sizes of 19 and 10 μm. Optical phonon lifetimes were determined by IR-V ASE. Additionally, we solved the linearized Boltzmann equation for phonons with first-principles force constants and performed molecular dynamics simulations.
The thermal resistance across the Al/AlN/sapphire structure barely changes for AlN thicknesses up to 30 nm, which is the ballistic regime. Beyond that, resistance grows with thickness: the transition to diffusion. Calculations of thermal conductivity accumulation versus phonon mean free path show that the contribution of phonons with paths shorter than 30 nm is negligible. The average group velocity of phonons in the ballistic regime is 1818 m/s, compared with about 7000 m/s in bulk AlN. Optical phonon lifetimes are short in the ballistic regime but increase by an order of magnitude upon the transition to diffusion and then remain nearly constant from 119 to 2440 nm.
This work provides the first direct experimental observation of ballistic heat transfer in AlN and determines the threshold thickness of about 30 nm. Compared with neighboring studies that looked at thermal conductivity or superconductivity in films, the key difference here is the systematic variation of thickness over a wide range and the separation of acoustic and optical phonon contributions. This refines the physical picture of heat transport in high-thermal-conductivity materials at the nanoscale.
The topic could develop toward engineering defects and boundaries to extend the ballistic window. Other substrates and interfaces could be explored, as well as the effect of doping. It would be interesting to test whether the ballistic regime persists at smaller thicknesses and how the quality of the AlN/sapphire interface affects it.
The results will influence the design of heat-spreading layers in gallium nitride transistors, ultraviolet LEDs, and high-power radio-frequency devices, where AlN is used as a buffer or substrate.
The next step is to measure the thermal conductivity of AlN on other substrates and with different defect densities, and to directly compare with first-principles calculations for specific microstructures.
For the thermal physics of thin films, the question of the ultimate ballistic resistance and the role of interface defects remains open. This work links microstructure (dislocations, strain) to thermal transport, but a precise quantitative description of defect scattering is still to be built.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
This study measured the thermal resistance of a series of AlN films on sapphire with thicknesses ranging from 1.6 to 2440 nm using the TDTR method. It was shown that for thicknesses below 30 nm, the resistance is nearly constant (ballistic regime), and then it increases with thickness (diffusive regime). Calculations revealed a decrease in the group velocity of phonons in the ballistic regime and an order-of-magnitude increase in the lifetime of optical phonons upon the transition to diffusion.
Where its strength is
Systematic variation of thickness over a wide range and direct experimental observation of the transition from ballistic to diffusive heat transfer in AlN, with separation of the contributions of acoustic and optical phonons.
Where it could go next
Worth trying
Why this may matter
Determining the threshold thickness for ballistic heat transfer in AlN could help in selecting the optimal thickness of heat-spreading layers in nanoelectronics, where controlling thermal resistance is critical.
Nearby works in our archive
The archive is sparse around this work: the closest review is only 0.59, against a usual 0.64. It looks like few have drilled nearby.
The 30 nm thickness at which the transition occurs is about 100 times smaller than the phonon mean free path in bulk AlN (about 3 μm). That is like a sprinter running a distance a hundred times shorter than his usual acceleration phase.