“Nothing in life is certain,” writes MIT mechanical engineer Seth Lloyd, “except death, taxes and the second law of thermodynamics.” But is this necessarily so? In episode 52, we’re joined by Andreas Schilling with the University of Zurich, who discusses his development of an amazingly simple device that allows heat to flow from a cold object to a warm one without an external power supply; a process that initially appears to contradict this fundamental law of physics. His open-access article “Heat flowing from cold to hot without external intervention by using a ‘thermal inductor’” was published with Xiaofu Zhang and Olaf Bossen on April 19, 2019 in the journal Science Advances.
Websites and other resources
- Olaf and Andreas’ original article on their thermal inductor
- University of Zurich press release (English, German)
Media and Press
Patrons of Parsing Science gain exclusive access to bonus clips from all our episodes and can also download mp3s of every individual episode.
🔊 Patrons can access bonus content here.
Hosts / Producers
Doug Leigh & Ryan Watkins
How to Cite
Leigh, D., Watkins, R., & Schilling, A.. (2019, June 25). Parsing Science – Bending the Laws of Physics. figshare. https://doi.org/10.6084/m9.figshare.8862209
What’s The Angle? by Shane Ivers
Andreas Schilling: I would say the more you know about physics, the less you believe that the whole thing is working.
Doug Leigh: This is Parsing Science. The unpublished stories behind the world’s most compelling science as told by the researchers themselves. I’m Doug Leigh. Ryan’s on vacation but he’ll be back next time. Today, in episode 52 of Parsing Science, we’re joined by Andreas Schilling from the University of Zurich who talks with us about his development of an amazingly simple device that allows heat to flow temporarily from a cold to a warm one without an external power supply, a process that initially appears to contradict a fundamental law of physics. Here’s Andreas Schilling.
Schilling: Hello! I’m Andreas Schilling. I’m living in Switzerland, in the beautiful city of Zurich. I was born actually in the countryside in Switzerland, but I went to a very good school, we had very good teachers there. I finally decided to study physics at the ETH in Zurich. So it’s quite a famous school in Switzerland. So I did my diploma there, and I was asked to do a PhD also, because my grades were not so bad so I was taken as a as a PhD student, and this was actually in superconductivity. So in that time, we discovered a new super conductor which is still high PSI at ambient pressure, so it’s a130 Kelvin also. And then people pressed me to go into academics, so I did a postdoc in Berkeley, so I went two years to the Department of Chemistry actually, but they did a lot of physics there. This was actually the best time of my life, so please ask me to go back there and probably say yes. And then after my contract ended I had to go back, and after that I came to Zurich also as a professor.
How & why Andreas carried out this research into thermodynamics
Leigh: Andreas primarily researches superconductors, which are substances that conduct electricity without resistance when below a critical temperature. So Ryan and I were curious to hear more about how and why he ended up carrying out this research into thermodynamics, the branch of physics which deals with the movement of heat between different objects, and its relation to energy, work, radiation, and the properties of matter.
Schilling: In reality, most of my measurements were based on thermodynamic properties, like measuring the heat capacity, so how much heat do you need to raise the temperature of an object by one degree or something like this. This is a quite a fundamental property for superconductors or for other interesting materials, there’re always theories describing the processes happening there, and of course, then there are predictions about these quantities, like heat capacity or magnetization or other things. So I did this also in Berkeley which was also very successful, and here in Zurich also I continued this line of research. And one of my last applications for the National Foundation, there was a student applying, which also came from the ETH and actually it was more like an electrical engineer type of person, I suspect that he was less interested in physics but more into electrical engineering and things like that, but that doesn’t matter, I mean if you’re good in the lab, if you can solder wires and things like that, this is fine, this is fantastic. Finally then turned out that he’s also good physicist but the fact that he was a good electrical engineer was very valuable, because at some point, he was down in the lab and he was doing something I didn’t really know what he was doing, and then they told me he wants to invent a new technique to measure heat capacity. And then I told him, well, this is very nice but this is not really the task of your PhD. I mean what you should do, you should measure physical quantities of superconductors, and how these electronic things in there are going on, and less invention of new techniques. But just go ahead and see what happens.
Leigh: Along with resistors and capacitors, inductors are one of the basic components of electronic circuits. They were independently discovered in the 1830s by Michael Faraday and Joseph Henry, whose name has been eponymously associated with the measurement unit used to gauge the degree of electrical inductance in a circuit. In 2011, Andreas and that student he was just describing, Olaf Bossen, published a study involving the use of an electrical inductor to effectively act as a thermal inductor. Eight years later, that work again became relevant to and raised his own primary research interests resulting in the article we’re discussing today. We asked Andreas to describe just what it was that Olaf invented.
Schilling: If you ever had a physics class in electrical components or if you even studied a semester in electrical engineering also, you may know that if you have a capacitor, which is a technical component, switch together with a coil, called an inductor, you get a circuit which oscillates, so it has oscillation electronic oscillations. And this oscillation frequency — so the number how many times it repeats itself — this frequency is determined essentially by the electrical capacity, and by this value of this inductance. So this is well known. And Olaf’s idea was to take the same idea, but to replace the electrical capacitor by a heat capacity. This was completely new; actually I didn’t believe it. He actually he did it, so he did it. So the trick was to translate temperature differences, which you need to measure heat capacities into voltage differences as for an electrical capacitor, and he did this by means of so-called Peltier elements, so this is a special electrical element that probably you know, but you don’t know the name of it, but you know it because if you go to a hotel room and you open the fridge, and usually these fridges, if the hotel is good enough, doesn’t make any noise, in contrast to your fridge at home where the compressor makes a lot of noise, but these fridges don’t make any noise, and the reason is because they have Peltier elements in there, so they are very special electrical components. So he took one of these, switched it in series with a conventional coil and this thing, in combination with a heat capacity, behaved like a thermal oscillating circuit. So instead of electricity, which is going plus and minus and plus and minus, you have heat flowing from right to left, and then from left to right, right to left, and so on and so forth. So this was his invention, and this was really fantastic.
The original box
Leigh: So a watch device was able to act like an alternating current, but instead of alternating the magnitude and direction of electricity, it did so with heat and cold. But it also required an external power source, making it an interesting novelty, but perhaps not so much a breakthrough in physics, as Andreas explains next.
Schilling: At that time, I, myself, already I thought well this electronic box you need to maintain this oscillation, is this really necessary? Because in the electrical world, when you have an electrical capacitor and an inductor, if the components are very expensive an ideal, so you have to spend a lot of money, then these this circuit can oscillate for a while without any external intervention. So it can oscillate for a couple of seconds, it can go up and down, and up and down in voltage, and of course, after a while it stops. And then I thought is that possible for heat, for our thermal currents, which are flowing in this device. Actually Olaf tried this, and it just made a lot of experiments, so he was soldering and soldering and soldering and trying, but it didn’t work, just didn’t work. So when you heat capacity up and you put it on top of this device, it just cooled down and nothing happened. And well, I thought that there must be some physics behind it to be considered, but I didn’t have time. So I just wrote down the equations, and then looked at them, and realized, well, it could be possible — maybe, maybe not. And then I put everything into a folder, and I just forgot it, and then a miracle occurred years later, and the miracle was that I obtained a sabbatical.
Energy required to power an autonomous circuit
Leigh: And so, while on a sabbatical in 2018, Andreas went back to investigate if the thermal circuit could continue to oscillate without the aid of Olaf’s electronic box. First, he crunched the numbers to figure out how much energy would be required to power such a circuit.
Schilling: So I was looking at these equations, and studying, and calculating, and then finally, I found the criterions that are necessary to have such an oscillation without any external thing, which drives the oscillation. Now this is one thing, the other thing is you want to have real numbers, you want to know what components do I really need in reality, and what is the oscillation frequency in blah blah blah. And this was a shock because you need an inductor, a coil, and the coil, the numbers that came out for the experts, it’s a 90 Henry, 90 Henry, this is impossible, I mean you cannot buy such a thing in an electronics shop, and 90 Henry call is a huge thing, it’s a several hundred kilos. If you want to minimize electrical losses, then you have to cool it down, in the ideal case it’s even superconducting, and that made me hesitant. So why shall I really go on with this because I will never be able to observe an effect? So this was a little bit of a shock. But then again, I remembered Olaf’s work. He also had to fight the same problem. In principle, I also needed very large inductances, and what he did because he didn’t care about energy consumption, he built some electronic circuit, which mimics an inductor, it’s just a replacement for a real coil, it has two inputs plus under — like a conventional coil — and his whole box behaves like a coil. Of course, it’s not a coil but it behaves like a coil, and this is the thing that he used for his experiment. And so I thought okay let’s do the same. First, we build such box, which behaves like an inductor on the reed, and we do the experiment.
Andreas’ device which acted as an inductor coil
Leigh: Having developed an electronic box which meant an inductor coil, Andreas carried out what would be the first of his experiments. Ryan and I were interested to learn how he did so, as well as what happened next.
Schilling: I, myself, as professor, I did the experiment. So you can imagine that in front of me there is a small copper thing which is quite heavy, this is the actual experiment. Some electronics around it, and then I had to heat this thing, and I just did it with a soldering iron — the things you’re using to solder wires together. So I just put it down and heated it up, and then my first try everything fell apart because it was too hot, but the second try worked and this was wonderful. So I had the time during that sabbatical to do this rather quickly and very carefully. So while I was convinced I had enough data to write a publication, I wrote a draft, already, that draft was really quite nice, I would say. What was not so nice was that we had to use these electronic books in the title of this paper. We claimed this is a thermal experiment without any external intervention and if you have to plug it in to get out some electricity, it’s not really without external insulation, but nevertheless, I tried it and, of course, it was rejected because of this. They said Okay, it’s nice, it’s very nice, but you have to prove it in an experiment, and justify this electronic box that we used. All questions they asked very easy, but this electronic box, this was really a thing that also caused me a headache, because it’s not elegant to have something that you have to plug in and claim it works without external power.
How the revised device fared
Leigh: Though he’d shown that isolation could be created, it still required this electronic box and that wasn’t quite good enough. This said, Andreas is off in search of a magnet that could generate 90 Henry of energy via a coil, but only facilities such as that at CERN in Geneva had such capacities. But then, he realized that he could use a superconducting magnet not as a magnet but as a coil. So he asked the senior researcher in his lab what kind of inductance might be obtained through the commercial superconducting magnets present in his own laboratory. The answer: 66 Henri; enough to get the job done. Andreas shares a story of what came at the second experiment after this short break.
Ad: We Share Science
Leigh: Here again is Andreas Schilling.
Schilling: What I decided to do I asked him to reserve my lab for myself. So usually my PhD students are working there and postdocs, so I told them I need half a day for this experiment, so please do your experiments in a way that I can have half a day for my experiment. And then, we did it. So I prepared my stuff, and put it on a small table, and put this table next to this magnet, and it just took me half a day to do it, and it just worked the first time I connected it, it just worked, and this was very good of course, because then I knew that the whole effect was possible. Now I didn’t tell you about the effect. The effect is actually in very short that heat is flowing from cold to hot without that you do anything from outside. I mean this is fantastic, because if you claim this to somebody the first time he will always say, especially if he’s a physicist, he will say this is not possible, this is simply not possible, because it seems to contradict physical laws: so heat cannot flow by itself from cold to heat. So what we saw, tell you first what we saw. So you heat something — something was a cube of copper maybe 10 grams or something — then you heat it 200 degrees Celsius, you place it on top of this device, and then you just release it, and see what happens. So there’s nothing connected to it, no electricity nothing, no electronics. The temperature goes down and down and down, as it should. So the expectation would be that the temperature approach is that of the underlying table or what it is. But what happened indeed was that the temperature was dropping down, below the temperature of the table. That means that at the moment when the temperatures were equal, this was cooling and cooling and cooling, and the heat was flowing from the chilling copper tube into the table which was warmer. So in other words, the heat was flowing from the cold object to the warm table, and this is something that you would never expect. So by using this electrical inductor in combination with this Peltier element, invented by Olaf, we created a circuit which can by its nature oscillate just by mathematics. So this is a mathematical oscillator, like a pendulum which also oscillates. It belongs to the same mathematical equations. So the fact that there is an inductance in there, a coil is really crucial. First of all it creates an alternating electrical current nobody is surprised, but as a consequence, because you have a Peltier element, you have also an alternate (16:59) at the same time.
Violation of the second law of thermodynamics?
Leigh: Like many others in the scientific media, at this point, Ryan and I were curious to hear how this effect doesn’t violate the second law of thermodynamics. Here’s what Andrea’s had to say.
Schilling: So the confusion is psychological. I would say the more you know about physics, the less you believe that the whole thing is working. And maybe at some point, if you really know a lot about physics, then you start to believe again. There are several laws of thermodynamics. I mean the first law is essentially energy conservation, so everybody knows it, oh almost everybody, so this is not disputed. For the second law of thermodynamics there are different versions of it, unfortunately, and usually, their physics teachers are celebrating these different versions. So they all having a list on the blackboard, and version one is: heat cannot flow by itself from cold to warm. The second one is that if you have a heat engine, its efficiency cannot be larger than a Carnot engine. Version 3 is the total entropy must increase with time, and bla bla bla. And then in the end, the claim is everything is equivalent, and I’m sure that none of these teachers know why it’s equivalent, because they’re just repeating what’s in the textbook. The most precise version of the second law of thermodynamics is not that heat cannot flow by itself from cold to hot, the most precise version is quantitative, and says that the total entropy of an isolated system must increase with time. This is useless for everyday, people, because they don’t know what entropy is, but for physicists it’s a very precise statement. So you can just calculate the entropy, and see what happens as a function of time. So let’s turn to the other version of this cold to hot, and hot to cold flowing. What is usually said is that heat cannot flow by itself from cold to hot, but this is not complete, it cannot happen unless there is some other change in the system going on, so something else must change at the same time. So this is the complete sentence. I mean the fact that heat is flowing from cold to hot is actually a very everyday experience in your refrigerator. Heat is always removed all the time from your coke that it keeps the low temperature, otherwise, it would warm up, and what his other change means here is that the compressor is working all the time, so something external is happening. So there are two ways of doing some change on the system: one is you would do external work for the compressors working, you plug it into your power cord, and it works. And there’s another way of other change, and that’s realized in our system. What happens is that the whole thing is, of course, also an electrical circuit, it’s not only a thermal circuit, it’s also an electrical circuit because it contains electrical components, like this Peltier element, it contains an electrical resistivity, because the wires have an electrical resistivity, it contains the coil. All these are electrical things, and there is an electrical current flowing flow. So what happens during this process is that you have an alternating current, and this current flows through the coil. We are using a small magnetic field, and as soon as the sign changes of the temperature and also of the voltage, the current changes, its direction and the magnetic field changes. Its direction and so on and so forth. So this other change which is going on at the same time, is the fact that there is an electrical current, which is alternating also, so it’s not only the thermal one, it’s also an electrical process which is going on at the same time. So this is it, I mean one has just get used to it. I mean there’s nothing else going on, it’s just this temperature oscillation, and this current oscillation at the same time, and that’s it. And as soon as the temperature of the body goes down below that of the table, then the heat is flowing from cold to hot. What is happening at the same time is that there is an alternating electrical current also flowing, but nothing else.
Potential applications of this technology
Leigh: It would seem that such devices tend to revolutionize everything from refrigeration and air conditioning to cooking and heating. So we asked Andreas to describe what some applications of this technology might be.
Schilling: One kind of application you could imagine is the following: you go to a desert, you have nothing else than your can of beer and our device, and this device doesn’t need any battery or something, so you can use it as often as you like. Okay so you go to this desert, and then you want to have a cold beer. So what you do, you take this kind of beer and put it into the sun, until the beer is almost boiling. Then you put your device into the shadow where it’s a little bit colder, and as soon as your beer is hot you put it onto this device and wait. And after a while, your beer is colder than its surroundings, ideally it’s 10 degrees Celsius, something like this. And then it’s cold, and then you can drink it, and you can repeat this as often as you like because it doesn’t consume any energy. So if you’re thinking of a large-scale application, what I would do if you had the electrical components that are ideal, you would have a hot object — maybe hot air — you want to cool down the room. You need a heat exchanger, and transfer the heat to the top part of this thing, and then you disconnect. And then everything cools down, and as soon as it’s cold enough below room temperature, you’re blowing out the cold air again, and then you have some cold in your room, and then you can repeat this a lot of times. And the nice thing is that the instance of its operation doesn’t require any energy. So principally, you can do this without consuming any energy which is in very contrast to air conditioners which use a lot of energy, of course. But I have to say this is in the very very far future, such ideal devices do not exist. So what I would need is a superconducting coil maybe, these Peltier elements actually are not very efficient, even those are used nowadays in high-power applications, they are not really efficient, and for this purpose, they would not be efficient enough. So there is some more research necessary in materials design. So you have to find new materials that produce a better effect, but this is not my business, I mean this is business of the material scientists, genius, you need a genius to invent.
Prerequisites necessary for breakthroughs in material science
Leigh: Our curiosity piqued, we followed up by asking Andreas what it might take to accomplish meaningful breakthroughs in materials design.
Schilling: Now I can tell you something about this, I mean this has nothing to do with physics. Innovation can happen in two ways: you can develop something gradually, step by step, and improving it gradually. I mean just a number increases step-by-step, every year a little bit — the cpu speed for instance — and this happens a lot in applications, and this also is the case for these Peltier elements. So maybe every two years there’s a very tiny step upwards, and then you with a new paper about a new material. The other way of innovation is that you discover something completely new, at once, which is ten times better. This also happens. So this happens for instance in the field of superconductivity, suddenly discover a new material with a critical temperature that means the temperature below which it’s superconducting, it’s only ten times larger than before. So this happens also, and my guess is that you have to wait for such an event. Instead of incremental improvements, you probably have to wait for a clever guy to find a nice material with better performance, but this could happen very fast.
Alternate means of operating the device
Leigh: Andreas made a short video of his experiment which is available at www.parsingscience.org/e52. In some applications of his device, temperature has to be either colder or hotter than ambient temperature of the environment. Since the temperature is oscillating, Ryan and I wanted to know how else this might be accomplished.
Schilling: I mean, what you could do, so this is one type how you could operate it: you could just wait half of an oscillation cycle. Usually an oscillation has several crucial time windows. So just think of a sign goes up, then it reaches maximum, then it goes down to zero, then it goes down to minimum, and then it comes back to zero. So the whole thing is one period, we call it one period. If you take only half of it, this is half a period. Let’s now talk about the cosine, which you may still know, instead of sine you take cosine, it’s thought that’s the maximum. Then it goes to zero, and after half of the period, it’s its minimum, and the same is happening here. It’s like a cosine starts as maximum temperature difference, and then it starts oscillating, going through zero, and going to the minimum. Now, instead of letting it oscillate all the time until everything is over, you just remove this connection when the temperature is smallest, and it will stay that forever if it’s isolated. So that means you could take can of boiling water connected with table temperature, let it down, and let it go to freezing, and then we move the connection, and it stays frozen, and you could repeat this every time, as many times as you want.
Leigh: Though most Parsing Science listeners won’t have access to a superconducting coil to replicate Andreas’s published experiment, we were eager to learn if listeners could duplicate the premise demonstrated by his and Olaf’s work.
Schilling: I must disappoint you in a way that average listener probably cannot do it, but who can do it: our college students, they can do it, or high school students, they can do it with the help of teachers. I mean, the good thing behind the whole concept of this invention is that it is with this electronic box, it is simple enough that if you have a technician at the school, who is good in soldiering and reading these diagrams for electrical circuits, it can be replicated. So I can send them — the scheme — for the electronics for this box, so you don’t have to buy a superconducting coil. So you build this box, the components would be around 20-30 dollars, and you would need a power supply to operate. It needs maybe five amps, so it should deliver, be able to deliver five amps. I don’t know how much this costs, but it cannot be the world, that the Peltier element itself is maybe two dollars, the copper around it is ten dollars, something like that, and we just need solder ring, and a little bit of glue. The other thing, you need a voltmeter. What I’m a little bit proud of is that if you are looking at the equations, and how the thing is connected together is essentially high school physics. A good student in Natural Sciences — maybe physics-related or engineering-interested — should be able to replicate everything, which is in the paper except for the actual experiment with the superconducting coil, of course. I claim that he could even understand the formulas, and see how these formulas work, and so what the meaning of the whole thing is. So this is just high school physics, or physics for engineering in the first year, and that’s why they can do it, they could really do it. So if you find a teacher who is also interested, I would be very willing to send them the blueprints of this electronic box, which mimics the inductor, and then you just have to invest some work, and you, of course, you need to voltmeters, and so on and so forth, but in principle, it’s very easy. The other thing is, I mean, the effect is so small with these components — it’s just two or three degrees Celsius — so it’s not really huge, but still, it’s you know, it’s measurable.
Links to article, bonus audio and other materials
Leigh: That was Andreas Schilling discussing his article: “Heat Flowing from Cold to Hot without External Intervention by Using a Thermal Inductor,” which he published with Xiaofu Zhang and Olaf and Bossen on April 19th, 2019 in the journal Science Advances. You’ll find a link to their open access paper at: www.parsingscience.org/e52 along with bonus audio and other materials we discussed during the episode.
Leigh: If you’ve enjoyed the first 52 episodes of Parsing Science, consider becoming a patron for as little as $1 a month. As a sign of our thanks, you’ll get access to hours of unreleased audio from all of our episodes so far, as well as the same for all of our future ones. You’ll help us continue to bring you the unpublished stories of researchers from around the globe, while supporting what we hope is one of your favorite science shows. If you’re interested in learning more, head over to: www.parsingscience.org/support for more information.
Preview of next episode
Leigh: Next time in episode 53 of Parsing Science, we’ll be joined by Bing Kim from Google Brain. She’ll discuss her research into classifying images based on high-level concepts using machine learning to assist both standard image classification, as well as in aiding experts with interpreting medical images.
Been Kim: My ultimate goal, when I started developing this method, was to care for lay people, lay people meaning someone who doesn’t have a computer science degree, why? Well, first reason is that you need doctors and legal experts to understand machine learning models, That’s when interpretability really matters. It’s not when you’re predicting movies what next to watch, it’s those high risk applications that interpretability will matter the most.
Leigh: We hope that you’ll join us again.