Rice University engineers taught microscopic silicon dots to “talk” to each other and accidentally found a better way to power the grid with wasted heat
Every year, the heat used to turn raw materials into consumer goods bleeds away before it can do useful work — costing the U.S. economy more than $200 billion in wasted energy annually.
At Rice University, engineers are attacking that problem from an unexpected direction. Rather than building bigger furnaces or better batteries, they designed a layer of microscopic silicon cylinders that “talk” to each other and pull photons, one by one, out of a bath of thermal radiation. The result is a heat-to-electricity conversion efficiency that conventional emitter designs have never been able to reach.
The enormous cost of wasted heat
Industrial heat loss isn’t a minor inefficiency — it’s a structural drain on the economy. Up to 20–50% of the heat used to transform raw materials into consumer goods escapes without doing useful work, costing the U.S. more than $200 billion every year. That number is hard to sit with.
The new emitter demonstrated efficiencies exceeding 60% while remaining application-ready — not a laboratory curiosity engineered for a press release.
Thermophotovoltaic systems — TPV for short — offer one of the most promising paths to recapturing that energy. They convert heat into light first, then turn that light into electricity using photovoltaic cells. TPV also supports grid-scale energy storage: unlike lithium-ion batteries, a TPV-based system could hold energy for long periods at potentially lower cost, a real advantage as grids absorb more intermittent renewable power.
The catch is that TPV systems have never quite lived up to their theoretical potential. The reason comes down to one underappreciated component.
Why thermal emitters have been the weak link
A TPV system has two jobs. The PV cell converts light into electricity — that part is well understood. The thermal emitter converts heat into light, and that part has received far less attention.
Conventional emitter design forces engineers into an uncomfortable choice: build something practical that performs poorly, or build something high-performing that’s too complex to deploy. Neither option makes TPV competitive at scale. Gururaj Naik, an associate professor of electrical and computer engineering at Rice University, identified this bottleneck directly.
“Using conventional design approaches limits thermal emitters’ design space,” he said, “and what you end up with is one of two scenarios: practical, low-performance devices or high-performance emitters that are hard to integrate in real-world applications.” His team set out to escape that trade-off entirely.
A quantum-physics playbook applied to heat
The breakthrough came from borrowing a framework more commonly applied to quantum systems. Rather than optimizing a single resonator in isolation, the team studied how an array of silicon nanocylinders behave when they interact — a concept rooted in non-Hermitian physics.
The emitter’s structure is layered: a tungsten metal base, a thin spacer material, and then the network of nanocylinders sitting on top. When heated, the base layers build up thermal radiation — a bath of photons waiting to be used.
The nanocylinders coordinate with each other, pulling photons selectively from that bath, and that collective behavior gives engineers precise control over both brightness and wavelength. “Instead of focusing on the performance of single-resonator systems, we instead took into account the way these resonators interact, which opened up new possibilities,” Naik explained.
Breaking the 60% efficiency barrier
That approach pushed performance higher than any practical design had previously achieved. The new emitter demonstrated efficiencies exceeding 60% while remaining application-ready — not a laboratory curiosity engineered for a press release.
Ciril Samuel Prasad, the study’s lead author and now a postdoctoral researcher at Oak Ridge National Laboratory, described the result plainly: the team showed how to reach the best possible emitter performance within realistic design constraints. The finding was published in npj Nanophotonics. Pushing beyond 60% would require new materials with better physical properties — not just more clever engineering of existing ones.
From factory floors to Mars rovers
The most immediate candidates are industries generating large, continuous streams of waste heat: nuclear power plants, steel mills, chemical processing facilities, and similar heavy industrial operations. At the grid level, the approach could underpin thermal-energy electrical storage — surplus renewable power stored as heat, then converted back to electricity on demand, where lithium-ion chemistry tends to struggle.
Naik also sees potential beyond Earth. Even a modest efficiency gain — from 2% to 5% — in rover power systems could represent a significant mission advantage. The research drew support from both the National Science Foundation and the U.S. Army Research Office. The next phase is expected to focus on pairing the new emitter with low-bandgap PV cells, the combination Naik believes could unlock the technology’s full promise.
Daniel García is an Editor-in-Chief with strong expertise in structural work and engineering principles. He combines this technical foundation with deep knowledge of energy, spatial design, and emerging technologies, bringing a forward-thinking and analytical approach to editorial leadership.
