21 March 2008--Engineers and scientists in Massachusetts have managed to greatly boost the efficiency of a common material used for thermoelectric cooling that has not been improved upon in 50 years. The researchers at Boston College and the Massachusetts Institute of Technology who reformulated the material--bismuth antimony telluride, or BiSbTe--say that not only will the change boost the efficiency of current uses but it will also open the way to operating automobile systems on waste heat from the engine and possibly provide an alternative to solar cells for converting the sun's energy to electricity.
Zhifeng Ren, a physicist at BC, and Gang Chen, a mechanical engineer at MIT, reported on their work in today's Science Express . They say that by breaking the bulk material into tiny chunks--from 5 to 50 nanometers across--they've increased a key measure of thermoelectric conversion, called the ZT of the alloy, from 1 to 1.4.
”This material has been used since the 1950s, and its ZT stayed at 1 for all this time,” says Chen. Ren adds that increasing the ZT from 1 to 1.4 is ”not marginal. It's huge.” The relationship between the ZT of a material and the conversion efficiency of a device based on it is not linear, so that translates into an improvement in thermoelectric conversion efficiency of between 15 percent and 30 percent, the researchers say.
They achieved the result using what Ren calls a ”very low-tech, very traditional technique,” called ball milling.” They placed a sample of BiSbTe into a shell and used ceramic balls to grind it into nanometer-scale bits. They took the resulting powder and placed it into graphite dies, in which they heated the material and applied pressure. The heat and pressure formed the powder into disks that were 1.25 centimeters in diameter and 2 millimeters thick, and into bars of 2 by 2 by 12 mm.
Not only did the ZT peak at 1.4 when the material was heated to 100 °C, it was also higher than the ZT of traditional BiSbTe across a range of temperatures. At room temperature, where the standard material has a ZT of 0.9 to 0.95, their nanostructured material reached 1.2. At 250 °C, their material had a ZT of 0.8, compared with only about 0.2 in the traditional material.
The reason for the improvement is that the nanoparticles create a high density of interfaces within the material. In a thermoelectric material, a temperature gradient across a material causes electrons to flow. The ZT can be improved by either raising the material's electrical conductivity for better current flow or by lowering the thermal conductivity to maintain the gradient. Heat moves across a material as vibrations called phonons, and in bulk BiSbTe, that movement is unobstructed. But in the nanostructured material, the phonons have to cross an interface every time they reach a new nanoparticle. ”Every time they see an interface, they see some reflection, and that reduces heat flow,” Chen says. Reducing the flow of heat increases ZT.
A number of researchers, Chen included, have also increased the ZT of BiSbTe and related alloys by building them into superlattices, in which an almost atomically thin layer of one material is laid down over another in an alternating pattern, creating numerous interfaces. Chen came to realize that a random arrangement of interfaces should work just as well as a periodic one, and that it would be easier than the ”very tedious” process used to create superlattices. It takes a week to make a superlattice 100 micrometers thick, the same thickness as the average human hair. With their new approach, he says, ”we can make tons.”
Most commercial applications of thermoelectric conversion are based on converting electricity into cooling power. TE coolers are used for temperature control in semiconductor diode lasers and some scientific measuring equipment. The largest consumer applications in the United States are probably for cooling car seats and plastic picnic baskets. Ren says TE coolers are popular in China for cooling drinking water, and for small refrigerators used in dorm rooms.
While those applications can become more efficient, Ren and Chen say the new material also opens up new applications, such as running air-conditioning systems, or taking the waste heat from a car engine and converting it into electricity to run equipment in the car. Converters made from the nanostructure BiSbTe would be efficient enough at converting ambient heat to electricity that they could compete with photovoltaic solar cells based on amorphous silicon at turning solar energy into electricity.
Peidong Yang, a chemistry professor at the University of California, Berkeley, who invented a silicon nanowire thermoelectric converter, calls the work ”quite interesting and significant, another great example of how defect engineering can significantly impact on the phonon transport in solids, and this time in a relatively cheap and easily processable materials system.” He says the ZT of 1.4 is high enough to make the material commercially interesting and adds, ”Given more improvement, it could open new opportunities in waste heat recovery.”
Ren and Chen say they plan to test their approach in a variety of materials to find a selection of alloys that can be used at a wide range of temperatures. They've formed a company, GMZ Energy, in Newton, Mass., to commercialize the material, which they expect will reach the market ”very soon.” Though they're not producing nanostructured BiSbTe in bulk yet, Ren jokes, ”If anyone wants to come and say, ’Sell me some material,' we might.”
About the Author
Neil Savage writes about lasers, LEDs, optoelectronics, and other technology from Lowell, Mass. For IEEE Spectrum Online, he wrote about solar cells in October 2007 and, in January 2008 about thermoelectric converters made from silicon nanowires.