Thermoelectric Heating Comes In From the Cold

Before architect César Martín Gómez could send his latest thermoelectric experiment to Antarctica in 2018, he had to make sure that soldiers from the Spanish Army could get it right on the first try. In the laboratory, he could always run the experiment—a scale model of a solid-state thermoelectric heater—a second time if it needed troubleshooting.

But at Spain’s Gabriel de Castilla base in the South Shetland Islands, soldiers would be too busy running other civilian experiments to troubleshoot Martín’s for him if it failed. And in a place like Antarctica, the goal of his experiment—providing efficient heat from direct current electricity—was both important and difficult. The experience “forced us to make a jump in quality,” recalls Martín, who is a professor at the University of Navarra in Pamplona, Spain.

That jump is part of a broader maturation of thermoelectric heating, and the demand for climate-friendly climate control technology is starting to grow. In September, the U.S. Climate Alliance, representing half the states in the United States, pledged to quadruple the number of heat pumps in the country by 2030. In November, the U.S. Department of Energy offered US $169 million in funding to help heat pump manufacturers expand. Heat pumps are one way forward, but thermoelectrics may represent a simpler next generation of climate control technology.

A decade ago, thermoelectric heating—which relies on the thermoelectric effect, in which electrons are pulled from one material to an adjacent material when both are heated—still required toxic or rare materials such as lead and tellurium. Because the thermoelectric effect is reversible, it also opens the door to cooling, with none of the environmental impact of using liquid refrigerants, or requiring industrial waste heat to be recovered as electricity.

However, the thermoelectric effect is so small in most materials and for small temperature differences that its real-world use so far has been mostly in space vehicles and for precisely controlling the temperature of donated organs for transplant. Researchers use a dimensionless quantity called ZT to describe the strength of the thermoelectric effect in any combination of materials. Two decades ago, combinations such as lead and tellurium yielded ZT values of around 1. After ten years, the search for new, more complex, and more effective materials had yielded ZT values of 2. In 2009, thermoelectrics researcher Cronin Vining wrote in Nature Materials that “commercial quantities of materials and/or efficient devices…does not seem imminent.”

But since then, materials scientists have been reporting more and more materials, such as tin selenide, and half a dozen other combinations, that lend themselves well to the thermoelectric effect. While some early startups covered by IEEE Spectrum have since gone under, one startup, Phononics, claims it is building the world’s first commercial building-scale thermoelectric climate control system. (Phononics did not answer IEEE Spectrum’s request for more information.)

Martín, meanwhile, says his technology, which his research group has been working on since 2009, is ready for commercial room-scale construction. For a typical 12-square-meter room, the cell would need to be 30 centimeters by 2 meters, and Martín envisions incorporating such a cell into the building’s facade. He says that would make it useful in homes, especially for renovations, because it doesn’t require tearing out walls or plumbing. “Our systems are much more expensive than a radiator, but the radiator has tubes behind the walls, a boiler, and a chimney, so it has more parts. The thermoelectric cell just goes in the facade and liberates the rest of the building, which architects love,” Martín says.

Advantages include next to no maintenance, because the thermoelectric cell has no moving parts, and avoiding the use of fluorinated refrigerant gases. It would also work well with solar power, because it works with direct current. 

Yet the technology is not ready for every setting: It’s even more expensive to use for cooling, although just how expensive it is also depends on the local climate.

Martín’s research group had partnered with a construction company starting in 2016, and that company was soon ready to incorporate the technology into their projects. The company’s management said they figured customers would like the ability to design their own floor plans, independent of radiator plumbing. But then various Spanish regional governments told the construction company that they considered the technology to be outside governmental building codes, so in 2018 the company stopped participating in Martín’s experiments. Instead, Martín thought to test his latest thermoelectric setup outside the reach of Spain’s building codes by sending it to the Gabriel de Castilla base.

While regulators catch up, researchers are fine-tuning how they purify easy-to-obtain materials such as tin selenide to make thermoelectric heating more cost-effective. Those researchers found a ZT of 3.1 for certain temperatures. Vining predicted in 2009 that a ZT of 4 would be the required threshold for commercialization.

Those results are just the open literature. When Martín sent a thermoelectric patent of his to a former commercial partner not long ago, the company stopped sending him new material on thermoelectrics. Thermoelectrics may be reaching the phase of industrial secrets and commercial applications. In other words, thermoelectrics may be mature enough to soon enter the market. 

For now, Martín and colleagues are sticking to test rooms and buildings. In Antartica, their scale model easily handled the extreme cold, corrosive marine air, and volcanic ash. The next iteration will have to take on building codes.

First published in IEEE Spectrum: [html] [pdf].