A semiconductor in an operating device can experience a strong electric field and considerable heat at the same time. Cornell researchers have investigated whether the field itself can change how heat moves through the material. Their work combines a computational study of aluminum nitride with separate measurements on hexagonal boron nitride.
The distinction between the two studies matters. One predicts how a field could affect thermal conductivity in a particular crystal. The other measures changes in thermal resistance in a different material and device structure. Together they suggest a useful question for electronics design, but they do not demonstrate a finished chip that cools itself.
The Cornell account published November 11, 2026 describes both projects. The experimental study appeared in Nano Letters on November 9. The researchers say further work is needed to understand and apply the observed behavior.
Why is heat a problem for electronics?
Electronic components move and control electrical signals, but some of the energy involved ends up as heat. If heat accumulates faster than it can leave, a device’s temperature rises. High temperatures can affect performance, efficiency and reliability.
The challenge can become more demanding when devices are small and powerful. A high power density means substantial energy is being handled in a limited space. Heat may need to cross thin material layers and interfaces before it reaches a surface where it can be removed.
Engineers use several strategies to manage this problem, including material choice, device layout and cooling systems. Understanding the properties of the materials while a device is operating is part of that work. A measurement made without an electric field may not capture every condition inside a working component.
How does heat move through a crystal?
Atoms in a solid are not perfectly motionless. Their collective vibrations can carry energy through a crystal. Physicists often describe these vibrational excitations as phonons.
Phonons can interact and scatter as they move. Those interactions affect how efficiently thermal energy travels. The material’s atomic structure, temperature, defects and boundaries can all influence the result.
An electric field can also affect the distribution of electrons in some materials. The Cornell team investigated whether that redistribution could change forces between atoms and, in turn, alter the behavior of the vibrations that carry heat.
That is a mechanism to be tested, not a suggestion that light, electricity and heat are interchangeable. Each has its own physical description, and a real device can contain several paths for energy to move.
What did the aluminum nitride model predict?
The first study focused on aluminum nitride, a semiconductor material of interest for high-power electronics. The team modeled a crystal between electrodes and considered electric fields applied in opposite directions.
According to Cornell, the calculations indicated that heat conduction across the material could increase or decrease depending on the direction of the field relative to the crystal’s polarity. The researchers attributed the difference to changes in electronic structure and atomic forces that affect phonons.
This is a directional result. It does not mean that applying any electric field to any semiconductor will improve cooling. The orientation of the material and the particular field conditions are central to the prediction.
A computational model can help identify a mechanism and suggest experiments. Its predictions still depend on the assumptions and approximations used, and a calculation for an idealized structure is not automatically the performance of a manufactured device.
What did the second study measure?
In separate work, the researchers examined thin flakes of hexagonal boron nitride placed between metal electrodes. They applied electric fields and used a laser-based measurement method to assess thermal resistance.
Thermal resistance describes opposition to heat flow. The measured resistance in this setup included the material and its interfaces with the electrodes. It should not be treated as a direct measurement of aluminum nitride’s thermal conductivity.
In one device, Cornell reports that thermal resistance decreased by as much as 25% during repeated tests. The changes developed over tens of minutes to hours, with recovery observed during or between measurement cycles. That figure describes a result in one experimental device, not a guaranteed reduction for every chip.
The team also found that conventional electrical measurements did not reveal the changes detected by the thermal measurements. That is a reminder that a device can appear unchanged according to one measurement while a different physical property is changing.
Conductivity and resistance are related, but not identical
Thermal conductivity describes how readily a material conducts heat under specified conditions. Thermal resistance describes how much a particular structure impedes heat flow. Geometry, thickness and interfaces can affect resistance as well as the material’s intrinsic properties.
Imagine heat crossing a thin layer between two contacts. The layer matters, but so does how well it meets the contacts. A change in the measured resistance of the whole arrangement cannot automatically be attributed to one component without further analysis.
Keeping those terms separate helps explain why the two Cornell studies are complementary rather than a simple model-and-confirmation pair. They involve different materials and different measurements. Each adds a piece of evidence about heat transport under electrical conditions.
Could this lead to cooler devices?
The aluminum nitride calculations suggest that engineers might consider crystal orientation when designing components that operate under strong fields. If a favorable orientation improves heat flow in a real device, that could be useful. The present result does not establish how much cooling a commercial product would gain.
Turning a laboratory finding into a design rule requires tests across operating temperatures, field strengths, materials and manufacturing variations. Engineers would also need to consider electrical performance, durability, cost and how the material interfaces with the rest of a device.
The gradual timescale reported in the boron nitride experiment raises additional questions about mechanisms and repeatability. A change that develops over minutes or hours may be useful in some contexts and less useful in others. Its long-term stability would need to be established.
What remains to be learned?
Researchers can test whether predictions hold in fabricated aluminum nitride structures and investigate which parts of the boron nitride device account for the observed resistance changes. Repeating measurements across devices would help establish how consistent the effect is.
They can also examine whether field-induced changes remain useful under realistic operating conditions. A device is a system of interacting materials, not a single isolated crystal, so a promising material-level effect must be evaluated in that wider context.
The work illustrates a useful shift in perspective: a material property measured in a quiet laboratory setting may not fully describe the same material under electrical stress. The next step is to determine when that difference can be predicted and used in practical electronics.
Explore the book: How to Do Hard Things: The Discomfort Trap
Be First to Comment