Researchers have introduced a high-performance thermoelectric material made from nanostructured antimony silicon telluride. The material, reported in the journal Joule, is not as efficient as state-of-the-art thermoelectrics, but still has exceptional efficiency, especially for a newcomer to the field, and could potentially be low cost and easy to make on a large scale.

“It has good enough efficiency to be taken seriously,” says Mercouri Kanatzidis, professor of chemistry and materials science and engineering at Northwestern University. “And people can work with it for the next few years.” Being a new material, it could take a while for researchers to better understand and optimize its properties for applications. Yet, Kanatzidis says, “the appeal here is to bring something new to the picture. It’s not taking something that’s already known and making it better.”  

Thermoelectrics can convert heat to electricity and vice versa, and have many potential applications. They can, for example, scavenge waste heat to generate power and efficiently cool structures. The best-known thermoelectric materials today are expensive bismuth telluride-based alloys. These and others such as lead selenide are already found in niche products such as climate-controlled car seats and in spacecraft power-generation systems.

Scientists have been searching for efficient thermoelectric materials that can be affordably made on a large scale. But it is hard to find materials with the right mix of properties. To be efficient—a measure called the thermoelectric figure of merit, ZT—thermoelectrics need high electrical conductivity and low thermal conductivity. But these two properties work against each other: improve on one and the other suffers.

Working intuitively, Kanatzidis and his colleagues tested various materials. They chose compounds based on a list of criteria: reasonable compositions without expensive elements; easy synthesis so it can be made in large enough amounts; semiconducting rather than metallic; and at least one or two heavy elements. “Most didn’t pan out, but this one did,” he says, speaking of the new Sb₂Si₂Te₂ compound.

The research team heated fine powders of Sb, Si, and Te, which they made in a ball mill, to make polycrystalline Sb₂Si₂Te₂ powder. Then they made pellets of this powder using spark plasma sintering, a process in which powders are pressed together while being hit with a pulsed electric current. The pellets yielded a ZT of about 1.08 at 823 K. A ZT of 1 is “seen as a value above which thermoelectrics can compete with other renewable energy systems,” says Matt Burton, a professor of materials engineering at Swansea University. A new thermoelectric material with a ZT over 1 is “exciting and noteworthy,” says.

The researchers went one step further. They treated the material in a liquid Te bath at high temperature, giving a material with a unique core–shell nanostructure: tiny crystals of Sb₂Si₂Te₂ with 5–20 nm ultrathin silicon telluride (Si₂Te₃) sheets at the boundaries between the grains so it looks like the crystals are wrapped in shells. These Si₂Te₃ shells act like filters, Kanatzidis says, allowing positively charged holes to pass through but blocking heat. As a result, the nanostructured material’s ZT jumped to 1.65.

One caveat of the material, says Burton, is the use of Te. Te is very rare in the Earth’s crust, with abundances similar to platinum, “so this material is not ideal for making large-scale cheap thermoelectric generators,” he says.

However, the core–shell strategy could be applied to other materials systems, says Jeffrey Snyder of Northwestern University, who was not involved in the work. “The nanostructures made here present a new type of grain boundary engineering in thermoelectric materials that could inspire the search for similar effects in other materials.”

Read the abstract in Joule.