Thermoelectric Paint Opens Prospect Of Easier Energy Harvesting

We will all be used to the thermoelectric effect in our electronic devices. The property of a junction of dissimilar conductors to either generate electricity from a difference in temperature (the Seebeck effect), or heating or cooling the junction (the Peltier effect). Every time we use a thermocouple or one of those mini beer fridges, we’re taking advantage of it.

Practical commercial thermoelectric arrays take the form of a grid of semiconductor junctions wired in series, with a cold side and a hot side. For a Peltier array the cold side drops in temperature and the hot side rises in response to applied electric current, while for a Seebeck array a current is generated in response to temperature difference between the two sides. They have several disadvantages though; they are not cheap, they are of a limited size, they can only be attached to flat surfaces, and they are only as good as their thermal bond can be made.

Researchers in Korea have produced an interesting development in this field that may offer significant improvements over the modules, they have published a paper describing a thermoelectric compound which can be painted on to a surface. The paint contains particles of bismuth telluride (Bi2Te3), and an energy density of up to 4mW per square centimetre is claimed.

This all sounds impressive, however as always there is a snag. The coating is painted on, but then it must be sintered at high temperature to form the final material. Then since the thermoelectric Seebeck effect voltage generated across a junction is tiny, some means must be arrived at to connect multiple regions of paint in series to achieve a usable voltage. The paint is produced in both n- and p-type semiconductor variants, so they appear to achieve this series connection by alternating bands of each. And finally the efficiency of the whole is only as good as the ability of its cold side to lose heat, so we are guessing to be effective it would require something extra to improve heat transfer away from it. Still, it will have a thermal bond with its substrate that is second to none and it has the potential to cover the entire surface of a hot item, so it shows considerable promise. The researchers discuss using it for power generation, but  we wonder whether there is also a prospect of it being used as a Peltier effect device to provide enhanced cooling.

We’ve covered many conventional thermoelectric generators in the past. The smallest was probably this LED ring, but we’ve also shown you a thermoelectric charger for use in rural Mongolia, and this very neat candle-powered fan.

Thanks to [Jack Laidlaw].

32 thoughts on “Thermoelectric Paint Opens Prospect Of Easier Energy Harvesting

  1. thank

    2016-12-09 22:00 GMT-08:00 Hackaday :

    > Jenny List posted: “We will all be used to the thermoelectric effect in > our electronic devices. The property of a junction of dissimilar conductors > to either generate electricity from a difference in temperature (the > Seebeck effect), or heating or cooling the junction (the P” >

    1. Probably too late by then. I had to look up Tellurium too. I’m more interested and worried in its relative scarcity. At least it’s a first step.

      *keeps reading*

      Nevermind, they’re used in rewriteable optical discs.

      1. Nah, if you impede the flow of exhaust gas, you’re ultimately taking power away from the engine. As for the heat of the exhaust, I can’t help thinking the same, that you’d lose more than you’d gain. Can’t think why, but it’s a general principle in things like this. Thermoelectrics work by slowing down heat transfer, they’re effectively thermal insulators that generate electricity.

        I know the exhaust doesn’t necessarily take away heat from the cylinder, so much as pressure. But would slowing heat transfer like this ultimately affect cylinder temperature (or put more load on the engine’s cooling pump otherwise)? It’s not obvious but it seems like there’s a hidden gotcha somewhere, that ultimately you’d be losing out, doing this.

        Maybe not. In which case it’s probably easy enough just to bracket a few thermocouples to a standard exhaust pipe. You don’t really need paint.

        1. you dont need to impede any flow with this though, just coat the inside of your exhaust system, well except for the catalytic converter of course, dont know if there is enough area to really produce anything tangible.

          no one claimed you needed paint, it just makes it a lot easier to get a very high surface area, ceramics are notoriously hard to bend…

          wouldn’t it be mainly waste heat?
          i really dont know, it might increase cylinder temp if it slows heath transfer through the exhaust, i dont know how much of the heat in an engine is exhausted and how much is cooled.

          1. Few things to bear in mind re automotives fall into:-
            & specifically re conventional Internal Combustion Engine vehicles up to prime movers
            1 – approx 65-70% heat exists via exhaust
            2 – most by radiative transfer along its length, ie prob 70% of that 65-70%
            3 – Thermals, flow & pressure intimately related by P1V1/T1=P2V2/T2
            So anything you can do to take heat away along the length of the exhaust
            pipe reduces (average) pressure which reduces pumping losses which means less
            piston pressure on the exhaust stroke which means greater engine efficiency.

            Selective heat removal might also affect tuning issues & affect torque curve a little…

            Its not clear if any thermoelectric devices (also how loaded re draw range) act as
            insulators overall or by their design improve conductivity – thats a whole series
            of curves significantly influenced by airflow around the car at various speeds.

            Casual surface area vs thermal flux vs exhaust temp differential calcs etc suggest
            a minimum worthwhile efficiency of around 10% or so for the paint or whatever provided
            it doesn’t add significant thermal inertia such as when standing in traffic jams – which
            might mean slightly more losses when less airflow idling reducing thermoelectric efficiency
            offset we hope by far greater gains when traveling at highway speeds…

            The static to dynamic inflection re cost vs Watts such as to displace need for
            the existing (very lossy) alternator at low speeds might suggest an impedance
            matching approach be taken to the whole thermo/alternator -> battery -> car
            electricals. In that respect I expect such a system would have more utility
            for a hybrid with some intermediate supercaps too which could also collect braking
            energy at same time.
            ie Between the low impedance braking to supercaps & ostensibly high impedance
            thermoelectric sources to the higher impedance typical Li-Ion batteries (which don’t
            like trickle charging) means a review of the whole thermodynamics. Could be an
            interesting issue to match up all the parameters. Eg Truck trials in Europe re these:-

            Friend of mine crafting a plan to use few of these & other power tech mixes for a 12 meter
            motor home which expects to achieve consumption of around 15L/100Km bio-diesel mostly.

            But, nonetheless thermoelectrics something that’s been considered by many for decades &
            lets hope material advances & ready access to engineering samples can offer enthusiastic
            well educated entrepreneurs & practical tinkerers opportunity to devise proof of concept
            systems of various sizes and means to collate their efforts.


          2. Heat recovery on a van or truck was done successfully and it did allow the removal of the alternator. It needed a lot of thermoelectric modules of the type with high melting solder. Recovery, IIRC was about 1kw for an overall efficiency of about 3%. I think it was a stunt by one of the peltier module manufacturers and I didn’t see full details.

          3. Don’t thermoelectrics, by definition, slow heat flow? Isn’t that basically how they work, or at least a necessary thermodynamic result of them working?

            An analogy would be an ordinary generator / motor, where the more current you draw, the harder it is to turn. It’s probably more complex than I can put into words, but that’s generally how energy conversions work, you get in the way of some kind of force (heat transfer being a force), and you get useful energy as a result. A completely efficient thermoelectric would surely be a perfect insulator.

            I’m struggling with the details, but I have a feel for this sort of thing, science and logic, born of experience. Most of us here probably have. I can usually tell when something’s off, or intuit it’s necessary principles, before I actually think it through. In this case I’ll have to go look up what actually happens in a thermoelectric to push the electrons round. I just bet it gets complicated really quickly.

          4. Think of a thermoelectric generator as a heat sink, just one that isn’t as good as copper. If it’s so good that the hot side temperature drops a lot then the power production is crippled. In an exhaust this isn’t really an issue. Nothing cares about cooling the exhaust, only getting rid of it. Cooling down the exhaust would be be the result of extracting energy from it, but it also means getting rid of 10’s or 100’s of kilowatts thermal energy to the environment and that may mean a lot of vents and airflow and horrible aerodynamics.

            If one is being put in an application that replaces a good heat sink, then a decent analogy might be putting a dam in the river.

        2. Going with that idea, who says exhaust tubes have to be completely round? Why not square or any other flat sided shape? To avoid the stagnant fluid (any medium that flows can be considered a fluid) in the sharp corners, radius the corners a small amount. Flat surfaces for the TE modules. Expand this idea further: Intake air through an outer tube to maximize the delta T, then cool the intake air in a intercooler before the engine intake. Intercoolers are used all the time on turbocharged engines, so why not?

          Am I crazy or would this help the efficiency of a hybrid vehicle?

          1. They don’t have to be round, but it is the most efficient. Square would have to be bigger to provide the same flow rate. This would cause it to use more material. This makes it heavier and more expensive for any given material. Whether this would totally negate other benefits would require data we don’t have.

        3. The thing that people are so quick to forget is that delta T is everything. There are many hot spots on a car with waste heat, most of the gas you burn does nothing more than make heat. The big problem with thermoelectrics is that the ‘cold’ side needs to be cooled, either by massive heatsinks or a liquid cooling system. Thermoelectrics are very inefficient so you won’t get a lot of power for all the weight you’ve added to the vehicle. A standard alternator is very efficient for the small amount of weight is adds. We don’t need thermoelectric paint, we need more efficient thermoelectrics.

  2. Can it make my pellet stove power itself for use during power outages? It’s just a useless hunk of steel with no power. Even if I had to pull start it like a lawn mower to get the fans turning..

    1. I could swear I saw a commercially produced in-flue TEG in the 300W range that might be adaptable to such a use, but can’t find it now. It was designed for wood/coal stoves, so of course you’d get reduced output with the lower flue gas temperatures of a pellet stove, but it might still be enough…

  3. in the future there will be a paint (some kind of nanotechnology) that will convert the infra heat spectrum to a visible or even uv radiation, they will put this paint all over the highways and to the cities (on the roof and on the pavement) so all the disturbing heat will be radiated to the space, the only drawback will be that at nigh we will have a very bright light, but that will be acceptable, i think

      1. astronomers can go to some desert locations, not to mention they even have a hard time right now with all of the city lights, as for the human sleep hormone, we can take that orally, like we do with the various vitamin supplement

    1. The cheap green laser pointers already do that. Using a non-linear crystal, whose operating principles are baffling. They use an IR laser to pump the crystal, which emits green laser light. The secret is basically, AIUI, a weird molecular structure.

      AIUI again, they’re not massively efficient, and how much IR does the Earth give out at night anyway? I think you’d be much better off, if you want to see, just using lamp posts. Or if you insist on using IR, a night-vision scope.

  4. The problem with this painted-on thing is it only works if you have hot and cold regions on the same surface. For the depicted hemisphere, with the pole hot and the equator cold, that’s great, but in real life, I’m not sure how often you would have that sort of temperature gradient on a single surface.

    It might be more useful if you’re driving it for the Peltier effect — if you painted this on a conical bowl made of some thermally insulating material, the outer portion (lots of surface area) would act as a heat sink, while the tip would heat/cool a small volume of liquid.

  5. I wonder if this would be made good use of on things like space rockets, satellites etc. Paint some of this over a part that faces the sun and the cold side in the shadows. Might be able to make extra power or use it as solar panel backup system.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.