Thermoelectric Milestone from UC Berkeley: Organic Thermoelectric Material
15 February 2007
 |
| A benzenedithiol molecule trapped between two gold surfaces. Click to enlarge. Credit: Ben Utley. |
Researchers at the University of California, Berkeley, have successfully generated electricity from heat by trapping organic molecules between metal nanoparticles to create an organic thermoelectric material. The new UC Berkeley study marks the first time the Seebeck effect has been measured in an organic molecule.
The discovery, described in a study published today in Science Express, could lead to the development of more cost-effective thermoelectric converters that could be applied to waste heat recovery—including in vehicles. (Earlier post.)
Utilizing wasted heat has been a major focus of research into thermoelectric converters, which rely upon the Seebeck effect, a phenomenon in which the application of heat to combinations of certain metals induces an electric current.
In 2005, the DOE selected BSST, a subsidiary of Amerigon, to lead the development of an efficient and practical thermoelectric system that will improve fuel economy by converting waste heat in automobile engine exhaust into electrical power. (Earlier post.)
Although the efficiency of thermoelectric materials has improved dramatically, it is still rather low and the materials are costly.
The goal is to make things out of materials that are more abundant and more easily processed. Organics are cheap and can be processed easily.
—Rachel Segalman, UC Berkeley professor of chemical engineering
The researchers coated two gold electrodes with molecules of benzenedithiol, dibezenedithiol or tribenzenedithiol, then heated one side to create a temperature differential. For each degree Celsius of difference, the researchers measured 8.7 microvolts of electricity for benzenedithiol, 12.9 microvolts for dibezenedithiol, and 14.2 microvolts for tribenzenedithiol. The maximum temperature differential tested was 30 degrees Celsius (54 degrees Fahrenheit).
The effect may seem quite small now, but this is a significant proof of concept, and the first step in organic molecular thermoelectricity. We are going down the road of cheap thermoelectric materials.
—Pramod Reddy, co-lead author
The next step for the researchers includes testing different organic molecules and metals, as well as fine tuning the assembly of the structure.
This research was supported by the US Department of Energy, the National Science Foundation and the Berkeley-ITRI Research Center. The Industrial Technology Research Institute, or ITRI, is a large research organization in Taiwan that is collaborating with UC Berkeley on nano-energy innovation.
Resources:
“Thermoelectricity in Molecular Junctions”; Pramod Reddy, Sung-Yeon Jang, Rachel Segalman, Arun Majumdar; Science DOI: 10.1126/science.1137149
Very Cool concept. If you could get this cheep enough then you could heat/cool your house using a heat pump where the temperature differential also generates enough electricity to run the pump.
Posted by: Neil | 15 February 2007 at 03:36 PM
Neil: a good observation.
Couldn't these just replace solar cells to generate electricity? Granted there must be a radiator mechanism somewhere to maintain a differential.
Posted by: K | 15 February 2007 at 03:53 PM
No, you could not have it generate enough electricity to run itself. It takes work to create the temperature differential for the heat pump to work. The greater the differential you try to create, the more work you have to put in. Also you need a greater temperature differential to create electricity. Therefore, when the thermoelectrics run at their most efficient is also when you would need the greatest amount of energy input to get any gains in cooling or heating.
Posted by: Patrick | 15 February 2007 at 03:58 PM
Patrick: A geothermal heat pump operates with a peak efficiency of about 400% using the existing heat differential between the ground and the air temp in your house. You don't require energy to create the differential you just use electricity to pump the heat one way or the other. You would need household electricity to start the process but once you have pumped enough of the heated/cooled liquid around in the system there would be enough energy in the differential to operate the system. The only question is whether or not you can harvest enough of it to keep the system going and still have much left for heating/cooling
http://www.bryantgeo.com/
Posted by: Neil | 15 February 2007 at 05:54 PM
Oh Boy!
Free heat...
Perpetual motion has been discovered.
Posted by: Lucas | 15 February 2007 at 06:11 PM
This may help with the problem of how to harvest low level heat which won't run a turbine.
Also if it can be scaled it could be used wherever there is friction or a chemical thermal reaction for waste heat harvesting.
Posted by: Ed | 15 February 2007 at 06:20 PM
har,har Lucas, you know very well there is no perpetual motion. I'm just talking about using some of the energy available in existing heat differentials.
Posted by: Neil | 15 February 2007 at 06:33 PM
What about a burner to act as a genset for an EV? Emissions should be pretty easy to control with a steady burn.
Posted by: John Ard | 15 February 2007 at 06:36 PM
Quick addition to my plan, you'd have to use the differential with outside air to create the power to run the pump or you'd lose power as the temperature of the inside air changed. The more extreme the weather, the more power the pump gets. If the ground temp was the same as the outside temp, you'd be dead in the water. So if you lived in a place where the average temperature was a comfortable one you wouldn't have to supplement the pump power. You'd have to run it as AC in the summer using outside heat energy to put some energy back into the ground.
Posted by: Neil | 15 February 2007 at 07:19 PM
To begin these thermoelectric devices have been around a long time. They are used for powering communications and controls for remote oilfield equipment. Low efficiency but reliable no moving parts to breakdown. The problem with these devices is their maximum efficiency is governed by Carnot's laws. For energy conversion it is 1-(Tl/Th). Where Tl is your heat sink lake water for cooling for example. Th is your heat source for example a black surface heated by the sun. Both temperature have to be expressed in kelvin or degrees from absolute 0 ( absolute 0 is -273 C or -460 F). Assume the heated surface is 50 C and the lake water is 10 C. the maximum efficiency one of the devices could have if it was perfect would be 1-((10+273)/(50+273))=12% efficient. But if the surface was heated to 350 C in a car engine the efficiency could rise to %55. Real devices are a lot worse than the Carnot ideal. Since discussion was talking a lot about heat pumps and refrigeration Carnot's law also governs them. The coefficient of performance determines how many units of energy can be moved from low to high or high to low with 1 unit of input energy between two temperature differentials(in a perfect device). COP for refrigeration is 1/((Th/Tl)-1) for heat pumps it's 1/(1-(Th/Tl)). So if you ran a heat pump to them power one of these devises the amount of power generated would only be a fraction of what you would need to run the heat pump. These devises have a uncertain future as a source of power in an automobile. If you added them to the engine they would insulate the engine to recover the waste heat. But insulating the engine itself would greatly reduce it's Carnot efficiency. So I am not sure if the net gain would be worth while. Maybe scavenging heat from the exhaust gases, but turbo chargers already do that.
Posted by: doug | 15 February 2007 at 09:59 PM
Neil: you completely lost me with your 'Quick addition...' comment.
I am always edgy when discussing heating and cooling. It is a tricky subject. But here goes:
I never thought you literally meant generating enough electricity to totally run the heat pump. Just some. And what is this about putting energy back into the ground?
Back to basics. The device generates a potential if one side is hotter than the other. Let side A be the air and side B be the heat exchanger of the heat pump.
When cooling your house side B will be hotter than A and there will be a voltage produced. But voltage is not power.
When a circuit is completed and current flows the device has an internal resistance. So it will heat up. Side A will warm and approach the same temperature as side B. As this happens the voltage drops and so must the power.
The device is self-limiting in the power it can produce. It will produce less on very hot days when you want airconditioning. And less on very cold days when you want heating. It can be useful but it won't be a panacea.
How does this differ from using it as a solar cell?
Sunlight striking the device can produce a high differential with the air even on the hottest days (sunlight heats the ground, not the air). So keep one side shaded and cooled by air while the other side is quite hot. Power can be produced in sunlight and at night when the air is cooler than the ground. The power limits would remain.
Now maybe we will hear from an expert. Patrick seems to approach this from thermodynamics. I use my hazy electronics.
Posted by: K | 15 February 2007 at 10:16 PM
Ok, what I had in mind was to take a regular geothermal heating/cooling system which consists of a closed circuit loop of fluid (moved by a pump) that runs through the ground and into the house. The fluid picks up heat from the ground and delivers it to a heat exchanger where the heat is removed from the fluid and released into the house. At maximum efficiency the amount of heat released into the house is 4 times the amount of heat you would get if you if used the electricity that drives the pump directly. In the summer you can turn the system arround and pump heat from the house into the ground.
To clarify, my idea is to add a station in the loop where the the fluid intersects with outside air. At this point some of the heat in the fluid could be used to generate electricity to help run the pump. If the efficiency of the generation is more than 25% then you come out ahead since that electricity will move more than 4 times the equivalent amount of heat. Obviously if you use too much of the heat from the fluid you won't have enough left to extract for the house and the efficiency will be reduced. In the summer the temperature gradient is reversed, heat is stored back into the ground.
With the help of Dougs equations I can see that the temperature gradient between the ground and the outside air wouldn't be anywhere near high enough to be able to generate electricity for the pumps and still have enough heat left to be extracted for heating. Wonder how many locations have their own hotsprings for the summer and glaciers for the summer?
Posted by: Neil | 15 February 2007 at 11:51 PM
One I am curious about is nuclear reactors, after they generate the electricity from steam and who knows, Hydrogen, there is still all the heat.
How much more efficient would a reactor become. Now add that to the some of the generation IV designs, makes you wonder.
Posted by: Paul | 16 February 2007 at 05:48 AM
Power plants, nuclear or fossil fuel, operate around 35% to 42% thermal efficiency so about 60% of the heat released is not converted into electricity. Many schemes have been considered to recover part of this "thermal polution" but as yet low temperature energy extraction remains a goal and not an achievement.
Posted by: Van | 16 February 2007 at 06:41 AM
Gas cogeneration reaches 60% efficiency (high temperature gas combustion turbine first, then steam driven turbines driven from the waste heat).
Posted by: clett | 16 February 2007 at 07:07 AM
Van,
I read once about the efficiency of a combined cycle power plant (gas turbine - steam turbine), and it was about 52% (overall).
Also, a big two-stroke (low speed) diesel generator can attain an efficiency of about 52% (mechanical)
Medium speed 4-stroke diesel generators can attain about 48.5% efficiency (mechanical).
Jorge.
Posted by: Jorge | 16 February 2007 at 07:12 AM
I was under the impression that nuclear plants were only ~5% efficient overall and that they produce massive amounts of waste heat. Could be wrong though....
I have always wondered why TECs were not used on the exhaust manifold of an ICE. Ex manifold temperatures in gasoline powered vehicles can reach 1,000 F. (600-700F more typical). That seems like plenty of good waste heat to me.
The ex manifold could come from the factory such that it is cast as a tube-in-tube design. The extra set of external tubes would function as a heat exchanger. Run liquid through this heat exchanger to a heatsink on one side of a TEC bank. The opposite side of the TEC bank could be cooled by liquid coming off of the radiator circuit.
Perhaps one could generate enough electicity off of such a design to charge the battery, run AC, and run all periphial electronics?
Posted by: John | 16 February 2007 at 07:48 AM
Doug,
Don't Carnot's laws apply to the harnessing of heat expansion? Thats not what is going on with thermoelectric devices.
Therefore, in (Tl/Th)..... your Tl is the expanded gas temp in a cylinder and your Th is your top dead center ignited gas temperature..... can't apply to a solid state heat device. Carnot's laws would not be applicable.
I think I've read somewhere that the theoretical efficiency is actually around 55%. Sorry, I could not even begin to locate that link but I believe that's what it was.
-JT
Posted by: Jay Tee | 16 February 2007 at 08:29 AM
Using TEC on the exhaust manifold isn't done for the same reason that there isn't a solar cell on the roof of most cars - the capital expense is too large relative to the electricity produced for the result to be profitable.
FWIW A few months ago I talked to a project engineer at a local electric utility. They were in the process of purchasing and installing several gas turbines. He indicated the efficiency was ~42%. I expressed shock at the low number. The combined cycle capital expenditure and maintenance to go up to 60%+ wasn't justified. If they'd wanted efficiency they would have gone diesel which would have been lower capital as well. I was more shocked at this, and asked why they'd gone to a turbine. He said that the production capacity per unit drove them to gas turbines: diesel gensets are too little.
Posted by: Harvey | 16 February 2007 at 08:33 AM
Jay Tee,
The Second Law of Thermodynamics is present in solid-state thermoelectric devices.
I once read in a magazine about a very efficient thermoelectric device capable of attaining an efficiency of 70% of the Carnot cycle.
Jorge.
Posted by: Jorge | 16 February 2007 at 09:12 AM
Harvey and clett,
I do not think a combined cycle power plant (gas turbine - steam turbine) can attain efficiencies of about 60%.
Where did you get this figure from?
Some years ago, I read an ABB advertising mentioning
a 52% overall efficiency for a combined cycle power plant.
Posted by: Jorge | 16 February 2007 at 09:26 AM
The thermal efficiency of LWR nuclear power plants runs between 35% and 42%. Ditto for big coal and big gas burning power plants. Yes, combined cycle plants do better, between 42% and 52%, and when you see numbers higher than that, the plants use the heat (space heating) and take credit in their thermal efficiency ads, but if we stick to electric generation, 52% is about it. Low temperature energy extraction remains a goal and not an achievement.
Posted by: Van | 16 February 2007 at 09:38 AM
Jorge: the 60% number for combined cycle comes from the General Electric "H System" CCG, model MS7001H/9001H. I'm not sure if they are taking advantage of space heating to improve their numbers.
http://www.ge.com/stories/en/20025.html?category=Products_Business
Posted by: Neil | 16 February 2007 at 10:00 AM
Neil,
after visiting the link you provided and also these:
http://cat.inist.fr/?aModele=afficheN&cpsidt=14198664
and
www.osti.gov/energycitations/product.biblio.jsp?osti_id=828052
it seems that they have already achieved 60% conversion efficiency from natural gas to electricity.
I am amazed!!
Thank you for the link.
Jorge.
Posted by: Jorge | 16 February 2007 at 11:24 AM
My knowledge of combined cycle plant efficiencies comes from more academic literature than sales brochures, so may need to be derated a bit. I've seen numbers as high as 63%. A brief internet search shows many references to 48-53% efficiency systems, as well as a few at 59-60%.
I was shocked at 42% because I know large (100,000HP) diesels can get above 50% efficiency converting heavy oil to mechanical power.
http://en.wikipedia.org/wiki/W%C3%A4rtsil%C3%A4-Sulzer_RTA96-C
Posted by: Harvey | 16 February 2007 at 12:24 PM