This is one of those stories that may seem a bit dense and technical at first, but should make you say "woah" (in your best Keanu Reeves voice) when the implications hit.
Two physicists -- Dr. Tammy Humphrey, Australian Research Council Fellow, and Dr. Heiner Linke, at the University of Oregon -- have determined that a particular structure and configuration of nanowires can have remarkable thermoelectric properties. Electricity can be generated from heat differentials across materials; historically, applications of this thermoelectric effect has been terribly inefficient, generally working at about 15% of maximum possible efficiency (the so-called Carnot limit). In a paper published in Physics Review Letters (PDF), Humphery and Linke have shown that specially structured nanomaterials can operate at much higher efficiency, perhaps even right up to the Carnot limit. What's more, the nanomaterial's thermoelectric effect is completely reversible, meaning that the application of electricity to the material would allow it to function as a heat-pump, pulling heat out of one end and pushing it to the other. The press release from the Nanoscale Device and System Integration conference (where the breakthrough was presented) is good for non-technical readers; the review of the article in Nature Materials online (free subscription required) is a bit more technical.
Thermoelectric generation is attractive for a number of reasons, including its utility at a variety of scales (from microscale on up) and its ability to take advantage of energy that would otherwise be wasted as lost heat. But the inefficiency of current thermoelectric technology limits its use. Current thermoelectric applications generally have a "ZT" rating (the measure of temperature-electricity conversion performance) of less than 2; a rating of around 5 is generally considered necessary for economical use. The Humphery-Linke model has a ZT rating of 10 at room temperature -- more than twice the level needed.
So what does this mean?
It could mean refrigeration without pumps and chemicals, and battery-powered "cold packs" able to maintain a given temperature for as long as the power held out (very useful for transporting sensitive medical supplies). It could mean high-end microchips able to operate at full speed with power-consuming fans replaced by heat-ferrying materials. It could mean very precise control over temperature for lab equipment and sensor technology.
It could also have significant applications in energy production and transportation. It could make marginal geothermal sites much more useful, opening up myriad new sources of non-polluting generation. It could be matched with photoelectric technology to capture additional electricity from otherwise wasted heat. And it has clear applications in hybrid vehicles, allowing for electricity to be generated from engine heat, providing an additional source for battery power beyond regenerative braking and direct engine recharge.
The Humphery-Linke article discusses only the physics of the idea, not the engineering. However, the Nature Materials review of the piece suggests that applications will not require new breakthroughs, and the thermoelectric nanomaterial model should be readily testable using "quantum dots." It's likely that real-world use won't match the maximum theoretical efficiency of the materials, but that's okay -- even if thermoelectric nanomaterial applications are only half as efficient as they could be, they could still be remarkably transformative.
One application you did not mention is solar power. Use simple mirrors or lenses to concentrate the sunlight, then use the nanostructured material to convert the heat into electricity.
My thoughts on thermoelectrics:
The concept has been around for a while and has applications already; increased efficiency will broaden its reach. Here are some of the advantages that I was able to pull together (in a transitional automotive-technology context):
* Simplification: Thermoelectrics can replace belt driven alternators and accessory systems (power steering, etc), as well vehicle heating and cooling systems, with a much simpler, principally solid state electronic system.
In the case of waste heat recovery, passing heat through a TE device generates current; similarly, passing current through a TE device can generate heat, or cold - reversing the polarity of current running through a thermoelectric system can change it from a heater to a cooler. This would allow for integrated, simple solid state, chemical free air conditioners and heaters, for instance. This makes for less expensive manufacturing, and less complex maintenance.
* Service Lifetime: TE devices have no moving parts, and no fluids or other materials that need period replenishment. Theoretically, they can run indefinitely, without maintenance of any kind.
* Waste heat recovery: This is the most obvious one - harnessing waste heat by building thermoelectric generation into vehicular systems. Traditional motors only convert 25% of the energy in gasoline to useable power; the DOE calculates that well implemented thermoelectrics could increase that to 50%. Excess energy can be pumped into batteries, or used to power electrical accessories or hybrid drivetrains.
* Production flexibility: Thermoelectric devices do not depend on physical orientation, belt drives, or fluid flow, allowing for much more flexibility and creativity in manufacturing.
* Improved underhood environment: Cooling off the underhood environment by capturing waste heat will also extend the service life of other vehicle systems, and increase their performance and efficiency. Take a look under your hood and note the number of heat sheilds that cloak your exhaust manifold and piping under the hood - this (and ceramic exhaust coatings) is manufacturer's current best effort at reducing temperatures.
This could be a big breakthrough in computer technology. If the theory is true very powerful computers such as graphic editing desktop machines could be manufactured in super small size. Maybe a third of the orignial size.
I enjoyed this article. Thanks
While looking for a project for my MEMS class, I happened upon the concept of thermoacoustic refrigeration. Very neat stuff, and it appears to be more-or-less what's being presented here.
The Fellows Research Group (Georgetown, TX) claims to have developed a MEMS device that uses the Stirling cycle to produce electricity or cool materials:
There's a link to a video (which takes forever to get around to the demonstrations) as well as a whole bunch of links to other research going on in the area (at the bottom).
Also, the Ben and Jerry's website has a nice little flash animation. I guess they've been funding a lot of the research into thermoacoustic refrigeration to cut down on their freezer emissions:
I wonder - could such TE devices be placed in attics of houses or even into the shingles to make electricity from the heat created from the solar effect?
I'm not sure if I understand TE equations all that well - does TE 'turn' heat into electricity such that in the end equation there's less heat and more electricity? If that's the case, then my idea previously would have the doubling effect of 'turning' the heat from the shingles and attic into electricity AND also keep the house cooler because less heat transfers into the house.
Thanks to Jamais for the posting the article.
In response to Erik's question, thermoelectric devices do 'turn' heat into electricity, and using them to generate power from waste heat will reduce the overall amount of waste heat that has to be removed - a tantalising thought when one considers applications such as computer processors, now putting out around 100W.
While this could be prospective application for very efficient thermoelectric devices (along with the recycling of the waste heat produced by car engines as pointed out by Rod), one has to keep in mind that the Carnot efficiency, 1-Tc/Th, gives the maximum fraction of the waste heat that you can recycle, which is quite small for small temperature differences such as that between your roof and your house. For this last application, solar cells (with an efficiency of above 10%) would do the best job.
If I've read the news bits right the new TE material could have a conversion efficiency of ~ 1/2 the Carnot limit. That means for a coal-fired furnace with ~ 1500 K fire and 500 K exhaust the Carnot limit is ~ 66.7%, and TEs could extract ~ 33.3%. Run the exhaust through another set of TE converters and exhaust at ~ 350 K, then you can pick up another ~ 15% and have a total 48% efficiency. Not bad. Making hydrogen from coal and using it in fuel cells could get about the same or a bit better.
There's an ethane-based heat-recovery system that ganged up to a coal-furnace with steam-turbines that can get a system efficiency of approaching ~ 60%.
What I want to really know is are TE materials able to handle the high heat differences needed to get good efficiencies or are TE applications forever condemned to recovering tiny fractions?
There's a new solid-oxide fuel-cell promising 50% efficiency using octane, but it runs at 800 degrees. With a TE extractor pulling energy from the waste heat in a hybrid car the system efficiency would be huge. A source at 1073 K and a sink at ~ 350 K means a TE can recover 33.7% of the waste heat. That's ~ 68% overall.
If the TE materials can hack the heat, that is.
Another application is using it in solar power satellites - imagine huge inflatable concentrators focussing light onto a TE converter running hot enough to get ~ 40% efficiency. That's as good as top of the line PV concentrator cells, and since ~ 60% is being ejected by a radiator the radiator efficiency should be good too (nice and hot.) If TE materials are cheaper than high-end PVs then such a system makes sense - even on Earth out in sunny climes.