The exhibit hall was buzzing like a high school science fair. At this year's Energy Innovation Summit, hosted by the U.S. Department of Energy's Advanced Research Projects Agency (ARPA-E), several hundred inventors and entrepreneurs had propped up posters to show off their bright ideas. At one table, a couple of Ph.D.s were grinning as if they'd already taken home a trophy. In a way, they had—their company, Phononic Devices, had made the most of a US $3 million ARPA-E grant it won in 2009, parlaying it into $10 million in venture capital money to commercialize its technology.
Phononic Devices, based in Raleigh, N.C., is building thermoelectric modules that use electricity to provide cooling power. Such modules essentially replace the flow of refrigerants through vapor compression cooling systems with the flow of electrons through a chip. CEO Anthony Atti says his company used the ARPA-E grant to "prove that we were right at the material physics level." Now, he says, the company is working on a cheap prototype that will "convince the market that we're right at the product level." The company's cooling modules are initially intended for electronics.
Atti's product will hardly be the first thermoelectric device to hit the market. The principles of the thermoelectric effect (which covers both cooling and heating) were discovered in the mid-19th century, and companies have found plenty of niche applications. Tiny thermoelectric devices are used to cool some CPU hot spots and laser diodes; shoppers can buy thermoelectric wine refrigerators and cars with thermoelectrically cooled seats.
So why is ARPA-E, an agency meant to encourage blue-sky energy ideas, investing in companies like Phononic Devices? Because so much more can be done to bring thermoelectrics into the mainstream, says ARPA-E director Arun Majumdar. "The compressor-based refrigerators and air conditioners are cheaper and have better performance today," says Majumdar, who worked on thermoelectrics at Lawrence Berkeley National Laboratory before taking charge of ARPA-E. The goal at ARPA-E is to improve thermoelectrics to the point where "the performance-cost ratio is not only comparable to the traditional ways of doing things but perhaps better."
Solid-state cooling would eliminate a system's moving parts and its refrigerants, which are usually potent greenhouse gases. And with the flip side of the thermoelectric effect, in which heat can be harvested to produce electricity, engineers could make use of waste heat in everything from car engines to factories. "Thermoelectrics could potentially be a game changer," says Majumdar.
Majumdar says both material and systems breakthroughs are needed before thermoelectrics goes mainstream. To create a cooling effect, voltage is applied to a compound semiconductor chip, which moves electrons from one surface toward the other. When an electron moves, it drags a phonon of heat to one side of the chip, where it can be dissipated by a heat sink.
Cooling chips are typically made of semiconductor materials like bismuth telluride, but these are woefully inefficient; today's thermoelectric devices often consume more than a watt of power per watt of heat moved. Researchers are struggling to develop new materials and structures that conduct electricity well but not heat—the needed combination to prevent transferred heat from immediately flowing back through the chip.
Phononic Devices' Atti says his company combines a new thermoelectric material with a thin-film design to form a cooling device that is two to three times as efficient as today's modules. He believes this efficiency level will make thermoelectrics commercially viable in mainstream applications.
Atti hopes to have prototypes ready for customers by early 2012. Other start-ups are on the same track. Tucson-based Tempronics says it has a different kind of cooling module structure that works with many types of materials.
Even if companies like Phononic Devices and Tempronics can put a thermoelectric fridge in every home, their work will be only half done. There's still all that waste heat waiting to be turned into electricity.
This article originally appeared in print as "A New Kind of Cool."