New treatment makes cheap thermoelectric material practical

Thermoelectric devices, which convert heat to electricity, may find broader use now that researchers have found a way to make a recently identified cheap thermoelectric material practical enough for real-world use (Nature Materials 2021, DOI: 10.1038/s41563-021-01064-6).

The devices could help harness the 65% of global energy produced that is lost as waste heat in places like smokestacks, car engines, and computers. The high cost and low efficiency of thermoelectric materials like bismuth-telluride-based alloys and lead selenide have hindered the materials’ widespread use, however, limiting them to niche applications like spacecraft and climate-controlled car seats.

Six years ago, researchers in the lab of Northwestern University chemist Mercouri Kanatzidis found that abundant, low-cost tin selenide is an excellent thermoelectric material. The catch is that it works best in its single crystal form, which is expensive to make and fractures easily, says Kanatzidis, so it’s “nice for research and understanding but not so practical for applications.”

Polycrystalline tin selenide is more robust, but efforts to make devices from it have met with hurdles. Heating and pressing tin selenide powder into polycrystalline pellets, which can be made into devices, leads to the thermoelectric performance dropping by more than 60%.

The reason for the drop is a thin layer of tin oxide that surrounds the small crystal grains in the polycrystalline material, Kanatzidis’ group found a few years ago. Tin oxide is a good heat conductor so it allows heat to travel between the grains, which is bad for thermoelectric performance.

Kanatzidis, In Chung of Seoul National University, and their colleagues have now found a simple two-step way to remove this oxide film. They first expose the oxide-containing raw material to hydrogen and heat, which removes the oxygen. After pressing the material into tin selenide pellets, the team treats the pellets again with hydrogen to remove any remaining tin oxide.

The resulting polycrystalline material performs more than 10% better than even single crystals.

Dmitri Talapin, a chemist at the University of Chicago, says these results are “truly impressive.” For the first time, he says, researchers have shown high performance for a material that checks every box for practical use: no expensive or rare components, no toxic elements, and made of polycrystalline powder which is cheaper, more scalable, and more mechanically robust than single crystals.

“It can be one of the most practically relevant recent works on thermoelectric materials for energy conversion,” he adds. “I believe this material can be commercialized on a relatively short time scale.”