In the pursuit of sustainable energy, some inventors think big. Zhiyu (Jerry) Hu, Ph.D., thinks small — at the nanoscale. His work in thermoelectric devices is transforming how we understand and harvest energy, turning minute temperature differences into reliable power sources.

For much of his life, Hu has found inspiration not just in fire but in heat itself. In 2005, while working as a physicist at Oak Ridge National Laboratory (ORNL), Hu captured global attention when he was featured in an Associated Press article holding burning glass wool adorned with platinum nanoparticles —so tiny they measured merely 10 microns in diameter, about one-tenth the width of a human hair. These nanoparticles, each only 10 to 20 nanometers wide, enabled what appeared to be spontaneous combustion without an external spark. Hu recalls the moment: “I have something in my hand that starts a fire without ignition. That looks dangerous, right?”

Fast forward to 2024, and Hu’s fascination with heat has been a theme that has led to advances in energy harvesting —one paper was published earlier this year in Nano Energy (2024, 123, 109393).  Hu and colleagues drew inspiration from the unique thermoregulation mechanisms of some beetle species, developing multi-bioinspired flexible thermal emitters. By mimicking the longicorn beetle’s fluffs structure and the Hercules beetle’s multi-phase interface, the scientists engineered a metamaterial that achieves robust cooling performance and energy generation capabilities.

More recently, Hu and his team introduced an approach to significantly enhance the performance of wearable thermoelectric generators (wTEGs). Their latest ScienceDirect paper, titled “Boosting self-powered wearable thermoelectric generator with solar absorber and radiative cooler” (Nano Energy, 2024, 132, 110381), discusses a novel method for integrating photothermal and radiative cooling techniques to maximize power generation in wearable devices. This approach allows wTEGs to achieve power densities up to 198 mW/m² for human body applications and 52 mW/m² for use with steel surfaces, such as robots, under practical outdoor conditions.

Reimagining energy

The Global Energy Challenge

The global energy crisis looms large over humanity’s future. Global electricity consumption is skyrocketing. The International Energy Agency (IEA) projects that global electricity demand will rise by 5,900 TWh in the Stated Policies Scenario (STEPS) by 2030, equivalent to adding the current level of demand in the United States and the European Union. The surge in interest in AI is also stoking more demand. Goldman Sachs projects that data center power demand could grow by 160% by 2030, with AI workloads a central contributor. Companies like Amazon and Google are planning to get their data center power from nuclear reactors.

Meanwhile, renewable energy sources like solar and wind have made significant strides in recent years but face limitations with intermittency, inefficiency, and infrastructure costs. Complicating matters is the fact that traditional power plants are inefficient. Roughly 60% of the energy in typical coal-fired facilities is waste heat. “You are throwing away two-thirds of the energy,” Hu said.

Making Δx small

The world needs fundamentally new and scalable approaches to energy generation and conservation. His early work with heat at the nanoscale has proved prescient. While others focused on generating more heat or creating larger temperature differences, Hu began thinking about heat differently. He recognized that traditional heat engines, based on principles established by French military engineer and physicist Nicolas Carnot, require significant temperature differences — often hundreds of degrees Celsius — to function efficiently. These engines are bulky and impractical for capturing low-grade waste heat or for applications at the micro or nanoscale.

In thermodynamics, Fourier’s law is often used to model heat flow where Q_cond = rate of heat conduction, k = thermal conductivity, A = cross-sectional area, ΔT = temperature gradient, and Δx is the thickness of the material or distance over which the temperature difference occurs.

Hu’s idea was to focus not on increasing the temperature difference (ΔT) but decreasing the distance over which that difference occurs (ΔX). “Carnot made the ΔT big; what I did is make the ΔX small,” he said. By minimizing the distance between the hot and cold sides to the micro and nanoscale, even a tiny temperature difference can create a significant thermal gradient, allowing for efficient energy conversion.

Potential energy sources, almost everywhere

It’s not hard to find temperature gradients. “Every surface has a temperature difference,” Hu said. “That temperature can turn into electricity,” Hu notes. The techniques he has helped develop open the door to harvesting energy from sources previously deemed too insignificant, such as the temperature variations between a person’s skin and the surrounding air. It also offers an energy source that operates continuously, independent of weather conditions or time of day — a core advantage over traditional renewables like wind and solar.

Colleagues have hailed Hu’s work as transformative. One noted in a supporting letter for the R&D Sustainability Innovator of the Year award that Hu’s research “represented a paradigm shift in the field,” enabling “efficient electrical power generation from very small temperature differentials as low as 0.001 K.”

Another remarked that Hu’s work “has revolutionized renewable energy solutions.”

Still, another expert highlighted that Hu’s technology is “particularly relevant for harnessing low-grade thermal energy, which is often wasted in industrial processes.”

Inventing new energy possibilities

While the ideas had been long simmering, maturing research findings into reality can sometimes be an enormous undertaking. For instance, Hu and his team worked on a specific thermoelectric device structure for eight years with thousands of precisely spaced air gaps. “I used about 4,000 wafers,” Hu recalls. The result was a device measuring just 2 cm × 2 cm × 6 μm that contains 10,082 thermoelectric P-N pairs. This third-generation chip can generate 2.6 volts at a 62.8 K temperature difference, achieving a power density of 0.5 W/cm³ and a power-to-weight ratio of 208 W/kg that are already comparable with traditional high-temperature burning combustion engines (Nano Energy, 2023, 114, 108611).

These devices exhibit the following performance metrics:

• Temperature Sensitivity: Capable of detecting temperature differences as small as 0.001 K, achieving two orders of magnitude higher sensitivity than existing technologies.
• Power Output Metrics: Outputting 2.6 volts at a 62.8 K temperature difference, with a power generation of 1.2 milliwatts. The devices achieve a power density of 0.5 W/cm³ and a power-to-weight ratio of 208 W/kg.

In 2024, Hu and his team introduced a novel method for sensing visual and auditory signals through thermal energy detection. Their micro-TEG chip functions similarly to a hardware Fast Fourier Transform (FFT) device, demonstrating high sensitivity in detecting the rate and intensity of input energy.

Scaling

Hu has a big vision for novel approaches to dealing with wasted energy. In a presentation, he predicts that the global waste heat recovery market will exceed $128.8 billion by 2030. But Hu sees even bigger possibilities in unexpected places — like greenhouses. In the same presentation, he projects that the market for temperature-differential power generation could reach $200 billion annually.

Hu’s commercialization strategy targets two distinct markets. The first involves energy harvesting for electronics,  especially “high-end instruments” used in laboratories. The second market is transformative in scale: converting waste heat into power across industrial and agricultural sectors. In power plants alone, these devices could significantly improve overall efficiency. The same principles apply to manufacturing facilities, where continuous processes in glass, ceramic, and steel production maintain stable temperature gradients that could be tapped for power generation.

The technology offers something distinct in the renewable energy landscape: consistency. Unlike solar panels limited by daylight hours or wind turbines dependent on weather conditions, these devices can operate continuously. “We can do like the leaves on the trees — to collect the energy quietly, continuously, every day,” Hu says.

“When we have electricity, basically, electricity is money already, right?” Hu notes. “If you know somebody who’d like to join our effort, I really welcome it because it’s a global effort. It’s not just in China, right? Everybody needs it.”