Controlling phonon dispersion will be crucial for improving how unwanted heat is removed from electronic devices, says team leader Alexander Balandin. Unwanted heat is a huge problem in electronics, and the problem gets worse as devices reach nanoscale dimensions. It is even a major stumbling block to miniaturizing components further.

Phonons affect all physical processes in solids, he explains. They limit how fast electrons move in materials near room temperature and impact on the optical properties of crystalline materials. Acoustic phonons are also the main heat carriers in electrical insulators and semiconductors.

Unsure about length scales

Similar to electrons, phonons are characterized by their energy dispersion relation (which tells us how the energy of a phonon depends on its wave vector, or momentum). Confining acoustic phonons in nanostructures affects how they propagate and modifies properties such as phonon velocity, polarization and density of states. It also changes the way phonons interact with other phonons, defects and electrons in a material. Confinement thus opens up opportunities for engineering the acoustic phonon spectrum in nanostructures and tuning the electrical and thermal properties of materials for particular applications, says Balandin.

Although predicted theoretically, modifying a phonon’s energy dispersion and velocity by spatially confining it has not been proven conclusively until now. What is more, researchers are still not sure about the length scale at which they can begin to change the phonon spectrum. Some believe that the nanostructure confining the phonons should be around a nanometre in size. Others think, however, that the phonon spectrum can be changed in larger structures as long as the confining structure is smaller than the phonon mean-free path (the distance a phonon travels before being scattered by other phonons).

Phonon spectrum change appears in nanowires as large as 120 nm

The new experiments by Balandin and colleges have established that the phonon properties start to change at length scales much larger than previously thought. Indeed, confined phonon polarization branches (which are proof of phonon spectrum change) appear in nanowires as large as 120 nm.

The researchers obtained their results by studying how phonons travelled through gallium arsenide semiconducting nanowires using an imaging technique called Brillouin–Mandelstam light scattering spectroscopy (BMS). BMS is similar to the more well known Raman spectroscopy, but measures light scattering on acoustic phonons rather than optical ones. However, since the energies of acoustic phonons are a thousand times lower than those of optical phonons, the hardware required for recording the signal is more complicated, says Balandin. Although the measurements are more difficult and time consuming, we were able to alter the energy spectrum, or dispersion, of the acoustic phonons we were studying by changing the size and shape of the GaAs wires they were confined in.

Engineering phonons could be the “next big revolution” in electronics

“Controlling the propagation of acoustic phonons in this way is key to when it comes to changing the way heat propagates in downscaled electronic devices, particularly at low temperature,” Balandin tells “It can impact the way we design electronic, thermoelectric and spintronic devices. For years, the only way to control heat propagation was to increase phonon scattering with nanostructured boundaries and interfaces. Our work shows that tuning phonon spectrum and phonon velocities is another versatile method for controlling thermal transport.”

Here, he makes an analogy with engineering electron wave functions by spatial confinement – which has become the foundation of modern electronics and optoelectronics. “Recent technological developments indicate that engineering phonons in this way will be the next big revolution, allowing for further increases in the integration density of computer chips and the efficiency of energy generation,” he explains.

Unwanted heat, and the difficulties in managing this heat, has become a major hindrance to further miniaturization of conventional integrated circuit technology, but researchers are now recognizing the fact that tuning phonon properties at the nanoscale will be crucial to solving this thermal bottleneck problem, he says.

The work, which was supported as part of the Spins and Heat in Nanoscale Electronic Systems (SHINES), an Energy Frontier Research Center funded by the US Department of Energy, is detailed in Nature Communications doi:10.1038/ncomms13400. Team members include researchers from Aalto University and the University of Eastern Finland.