Phononic Devices’s Chips Convert Waste Heat into Electricity
In 2008, the Rockefeller Family’s Silicon Valley venture capital firm, Venrock, tapped chemist-turned- venture-capitalist Anthony Atti to evaluate promising thermoelectric research conducted by Patrick McCann, a professor of engineering at the University of Oklahoma. McCann’s technology showed the potential to use semiconductors to capture waste heat and convert it into power, as well as to displace heat to cool everything from personal computers to tractor-trailer refrigeration units.
About 55 percent of all energy consumed in the U.S. returns to the environment as wasted heat, according to energy systems analyst A.J. Simon of the Energy Dept.’s Lawrence Livermore National Laboratory. Traditional thermoelectric devices used for cooling are quiet, compact, and reliable, but efficiency has been very low since they were first commercialized more than 25 years ago. McCann’s breakthrough could make them more efficient by using new materials.
In 2009, with $1 million from Venrock and Oak Investment Partners, Atti and McCann exclusively licensed the technology from the university, launching Phononic Devices in Raleigh, N.C., with Atti as chief executive officer. Today, the 16- employee company makes thermoelectric semiconductor chips that capture waste heat and convert it into usable electric power or -- depending on the source of heat -- provide refrigeration and cooling.
Phononic Devices has received a total of $12 million in venture capital from Venrock and Oak, plus $3 million from the Energy Dept.’s Advanced Research Projects Agency Energy program. Atti expects to start selling the devices to electronics cooling-and- refrigeration customers near the end of 2012. Phononic Devices is one of just seven thermoelectric device makers specializing in energy harvesting and cooling.
Atti, a 37-year-old organic chemistry PhD, spoke recently to Bloomberg.com contributor Karen A. Frenkel about where Phononic Devices fits in the $25 billion waste-heat-recovery industry. Edited excerpts of their conversations follow.
Karen A. Frenkel: What is the origin of your company’s name?
Anthony Atti: A ‘phonon’ is a particle of heat -- a thermal sound wave that vibrates through a material. An effective thermoelectric material insulates against heat by deflecting phonons (heat) and causing the phonon to lose its thermal energy. An ineffective thermoelectric material absorbs phonons and heats up itself.
Q: What was unique about what you saw in Professor McCann’s lab that made you recognize its promise?
A: For a thermoelectric material to effectively manage heat, it must do two things really well at the fundamental physics level: It has to be electronically conductive, yet thermally insulating. Those two processes have been very difficult to decouple. McCann presented experiments to do that with semiconductors that have not been explored for this purpose so far. The icing on the cake from the investment and economic perspective was that we could manufacture the material in a high-volume, low-cost manufacturing manner at the early stage in the company’s life.
Q: Why hasn’t the thermoelectric industry already tried these other semiconductors?
A: They’ve been hiding in plain sight. They’re used for other applications like optics [and] lasers, and other areas of the semiconductor world use them successfully -- but not for thermal management.
Q: What does the material do and how is it different or better from what’s out there?
A: We demonstrated that the material decouples electronic conductivity from thermal insulation. It manipulates the direction of electrons at the nanoscale.
Q: How does decoupling help harness waste heat?
A: Our fundamental advance allows us to deliver devices that can provide cooling for refrigeration or waste heat recovery and efficiently convert it into power. It can compete head-to-head against the incumbent technology, which uses the elements bismuth and telluride.
Q: How efficient is your chip?
A: Our materials and devices are expected to more than double thermal-electric efficiency -- compared to conventional thermoelectrics -- for the interval between room temperature, which is 73F, and 248F.
Q: How is your device different from what others in the semiconductor waste-heat and cooling area are doing?
A: There is definitely competition in the electronics cooling area and also in high-temperature waste-heat recovery for automotive or power plants. There is also competition in very low-temperature waste-heat recovery -- sensors, detectors, and wearable materials for the military. There are “pure plays” at the very low temperature range or at the high end. We believe that’s because operating within that low-grade temperature sector is really hard to do. But we have the three legs: performance (the fundamental ability to decouple thermal and electronic mechanisms), cost (efficiency with which we can manage heat), and manufacturability (using existing semiconductor processing in our path to market).
Q: What’s the size of the markets you’re going after?
A: The global electronics cooling market for applications between 70F and 250F is $4.5 billion, refrigeration is $6.5 billion, and harvesting low-grade waste heat is about $3 billion. The thermoelectric-specific market -- that is, of modules made from semiconductor materials engineered for thermoelectric behavior -- is $300 million in global sales annually.
Q: Explain your business model.
A: We’ve transitioned from proof of concept to proof of chip and are now moving into device fabrication. We are packaging the material into a high-efficiency-semiconductor thermoelectric device and will sell it to end users in cooling, refrigeration, and waste-heat recovery for power. That’s the business model today, based on the low-cost, high-volume tools available.
Q: What is the market opportunity?
A: Heat is a major problem that impacts almost every segment of the economy, but not all heat is created equal. Incumbent technologies can handle heat greater than 250F, which is the power generation market, pretty well. Converting waste heat to power is done typically with heat exchanges or industrial boilers. Some turn it into power, some use it elsewhere in the plant to heat up equipment, [and] some pump it out of a smoke stack. It gets better the hotter you go. But below 250F -- low- grade heat waste -- the incumbents don’t work that efficiently. Our technology works very well between 70F and 250F. That’s how Phononic is able to compete. Technologies that serve those markets now don’t do it very well because heat in that temperature range is notoriously difficult to deal with.
A: Because there is not a lot of it. Whether a compressor, heat exchanger, or thermoelectric device, your efficacy is dictated by the temperature difference you can maintain from a hot to a cold side. You need to move from hot to cold to get rid of it, whether cooling or capturing heat for power. If that temperature gradient is small because you’re operating at low temperatures, efficiency is really challenging. Our data show we’re really good at that temperature range.
To contact the reporter on this story: Karen A. Frenkel at firstname.lastname@example.org
To contact the editor responsible for this story: Nick Leiber at email@example.com