As electronic devices become smaller and more powerful, keeping them from overheating is an increasingly difficult engineering challenge. Now, Cornell researchers have developed a computational method, backed by the first experimental measurements of its kind, showing how an electric field can change the way a semiconductor conducts heat.
The research, led by Zhiting Tian, the John Brancaccio Professor in the Sibley School of Mechanical and Aerospace Engineering, predicts that the cross-plane thermal conductivity of aluminum nitride, a promising material for next-generation electronics, can either increase or decrease depending on the direction of an applied electric field.
The findings, described in the paper “Thermal Conductivity of Wurtzite Aluminum Nitride Under External Electric Field,” in npj Computational Materials, suggest that engineers could eventually design electronic components that dissipate heat more effectively under an operating electric field by aligning the field direction favorably with the crystal polarity.
“When we think about the thermal conductivity of a material, the way we usually model it is without an applied electric field,” Tian said. “But in real device operations, especially for those of high power density devices, the electric field is actually significant.”
The Cornell team focused on aluminum nitride, an ultrawide-bandgap semiconductor capable of withstanding substantially higher electric fields than conventional semiconductor materials. Such materials could enable smaller, more powerful electronics, including high-voltage power systems and high-power radio-frequency devices. But greater power density also produces more heat, potentially reducing efficiency, reliability and device lifespan.
Tian and her students, including first author and doctoral student Joyce Christiansen-Salameh, suspected that electric fields might influence the atomic vibrations responsible for transporting heat through these materials.
Those vibrations, called phonons, carry thermal energy through a crystal. As phonons travel, they interact and scatter, affecting how efficiently heat moves through the material. The researchers reasoned that an electric field could redistribute electrons within the crystal, changing the forces between atoms and, consequently, the behavior of phonons.
To investigate, the team developed a computational framework combining density functional theory (DFT), which describes the electronic structure of materials, with a technique called nonequilibrium Green’s function (NEGF). The combined approach, known as DFT-NEGF, had previously been used to study electron transport but had not been applied in this way to calculate phonon-based thermal conductivity under an electric field.
The researchers simulated aluminum nitride crystals positioned between electrodes and subjected them to electric fields in opposite directions. They calculated how the fields redistributed electrons, changed the forces between atoms and altered the scattering of phonons.
The results showed that the electric field could strengthen or weaken certain atomic bonds, depending on its direction relative to the crystal’s built-in polarity. Those changes, in turn, influenced how readily heat traveled through the material.
With no electric field applied, the calculated thermal conductivity perpendicular to the aluminum nitride film was approximately 334 watts per meter-kelvin. Applying an electric field in one direction increased that conductivity, while reversing the field decreased it.
The effect suggests a new consideration for designing high-power electronics. Rather than treating thermal conductivity as a fixed property, engineers may need to account for how it changes during operation. More intriguingly, they might deliberately orient semiconductor films so that operating electric fields enhance heat dissipation.
“We could intentionally grow the film in the direction where the electric field is favorable to help dissipate heat,” Tian said.
Putting theory to the test
A complementary experimental study led by Tian moves beyond predictions to measurements.
In research published Nov. 9 in Nano Letters, the Cornell team investigated hexagonal boron nitride, an ultrawide-bandgap material widely used as a dielectric in emerging electronic devices. While the aluminum nitride study developed a computational framework for predicting changes in heat conduction, the boron nitride study examined the phenomenon experimentally.
The researchers placed thin, single-crystal flakes of boron nitride between metal electrodes, applied strong electric fields, and used frequency-domain thermoreflectance, a laser-based technique, to measure changes in thermal resistance, a measure of how strongly the flake and its interfaces with electrodes impede the flow of heat.
In one device, thermal resistance decreased by as much as 25% during repeated tests. The changes developed gradually over periods of tens of minutes to hours, with recovery observed during or between measurement cycles. The researchers also found that conventional electrical measurements did not reveal the changes detected by their thermal measurements.
The findings, reported in “Reversible Electric-Field-Induced Changes in the Thermal Resistance of Hexagonal Boron Nitride,” establish an experimental approach to probing electric-field effects on thermal transport and provide experimental evidence that heat transport in an ultra-wide-bandgap material can change substantially under electrical stress.
The first author of the study is doctoral student Amelia Schaeffer. Co-authors include Lena Fine ’25 and Aloyse Maille ’27, both of whom were Undergraduate Research Grant awardees from the Duffield Engineering Office of Inclusive Excellence. This study was conducted in collaboration with James H. Edgar, professor at Kansas State University.
Both studies were sponsored by the Office of Naval Research.
