Engineers have come up with a handful of uses for computer chip-like devices that chill objects when plugged in or convert waste heat into electrical power—stuff like car seats that cool drivers on hot days and coolers that chill drinks when plugged in. But by-and-large, these devices, known as thermoelectrics, have remained too inefficient to make much of a real-world impact. Now, researchers in the United States and China report that they've come up with a new way to boost the performance of one of the most common thermoelectrics on the market , an advance that could pave the way for more widespread use in converting waste heat from cars and other mechanical devices into useful electricity.
Although efforts to improve thermoelectrics haven't paid off in a big way, it's not for want of trying. Physicists realized in the early decades of the 1800s that heat flowing in a circuit between two different conductors could generate an electric voltage. They also discovered the opposite effect -- electricity fed into such a device would heat one conductor and cool the other. The devices work because heat can push electrons around, and the motion of electrons can carry heat. Researchers have long tried to enhance the effect to make the devices practical. In doing so, the goal is typically to increase a property in the materials known as ZT, which depends on a set of factors that include a material's ability to conduct heat and its electrical conductivity. An alloy of lead telluride (PbTe), for example, which has long been used to generate electricity aboard satellites, has a ZT of around 0.8.
To increase ZT, researchers typically try to increase a material's electrical conductivity as much as possible while holding down its thermal conductivity. In 2008, researchers led by Jeffrey Snyder, a materials scientist at the California Institute of Technology in Pasadena, spiked PbTe with thallium , which boosted the ZT to 1.5. The group later determined that the thallium altered the electronic structure of the crystal, improving its electrical conductivity.
But thallium is toxic, so Snyder and his colleagues wanted to determine if they could match the improvement with other additives. Earlier this year, Snyder and his team at Caltech reported  in Energy & Environmental Science that substituting sodium for thallium produced a ZT of 1.4. Now, Snyder's team, in combination with researchers from the Chinese Academy of Sciences' Shanghai Institute of Ceramics, report online today in Nature that adding selenium and sodium gives them a maximum ZT of 1.8. The selenium not only further improves the electrical conductivity, it also reduces the thermal conductivity, Snyder explains.
"It's excellent work," says Gang Chen, a thermoelectrics expert at the Massachusetts Institute of Technology in Cambridge. "It shows there's still room to improve existing materials," he says Snyder notes that the same strategy should improve the electrical conductivity, and thus the ZT, of other conventional thermoelectrics. With any luck, the improvements will be large enough to push thermoelectric devices out of niche applications and into the mainstream.