Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Maxwell’s Demon - An Energy Conversion Experiment in Contradiction to the Second Law

  1. Jun 1, 2004 #1
    Maxwell’s Demon
    An Energy Conversion Experiment in Contradiction to the Second Law​

    We have performed successfully an experiment of energy conversion of a totally new type. Please see the attachment figures. The experiment relates to a special electronic tube, which has two similar and parallel Ag-O-Cs cathodes, A and B, lying on the surface of an insulated base. A and B are very close but well insulated each other, and they eject thermal electrons at room temperature ceaselessly. Over the narrow interval between A and B, the two cathodes may exchange thermal electrons. As the flow of electrons from A to B equals the one from B to A, no net accumulation of electrons occurs in either side. If we now bring a uniform magnetic field to bear on the electronic tube, the thermal electrons will fly only from one cathode to the other, resulting in an electric potential difference between A and B. The potential difference slows down the follow-up electrons as they fly passing over the interval. The electrons are cold down (slightly), causing the whole tube to cold down, too, (much more slightly). Hence the tube will extract heat from exterior to compensate its change in temperature. Accompanied by these processes, there is simultaneously a conversion of energy from the heat extracted from the surrounding air to electric energy of a direct current supplying to an outside load. The maximum current got in the present experiment is 8.0×10-13A, corresponding to a single way flow of electrons from one cathode to the other at a rate of five million electrons a second. Compared to Maxwell’s original idea of an intellectual being, which was supposed capable to interfere or control the molecules’ random motions in a gas, the magnetic field here is really a very skillful demon.

    For more details, please refer to our paper Realization of Maxwell’s Demon. It is on the webpage of http://www.arxiv.org/list/physics/0311?100 , No. physics/0311104. And you may download the paper directly from the web address http://www.arxiv.org/pdf/physics/0311104. We also enclose the paper as one of the attachment. Please check.

  2. jcsd
  3. Jun 3, 2004 #2


    User Avatar
    Science Advisor
    Gold Member

    Pretty interesting paper, well presented.

    1. I did notice the current produced was VERY low, and at the bottom end of the range for the Chinese-manufactured test equipment:

    "The current measuring range of ZC43 (produced by Shanghai Sixth Electric Meter Factory) is from 1 x 10-5A to 1 x 10-14A."

    2. You would have to think that the output current (8.0×10-13A) is less than what it would take to maintain a magnetic field of equivalent strength to the magnets used in the experiment. I mean, the magnets came from somewhere and it took energy to create them. That isn't free... is it?
  4. Jun 3, 2004 #3


    User Avatar
    Science Advisor

    I find it interesting too. To help me understand, I thought of the following analogy. Holding a pool of liquid like water or gasoline in your hand, you notice it evaporates: molecules 'jump' into the atmosphere and do not return. In consequence, the liquid cools. So in this experiment, electrons 'evaporate' and the cathode emitting them is cooled. No mystery there. The real trick is of course to "close" the system. Find a way to get the water molecules back into the pool that costs no energy. THen you would have a free fridge. Of course that is impossible. In the experiment discussed in the paper, electrons travel through an ammeter, completing the circuit. This circuit would be "closed" if the ammeter were entirely passive. But I doubt that is the case. I guess it needs electrical power to operate: Power that is absolutely humongous compared with the tiny voltage and current in the circuit. So for me the jury is out until I know much more about how the meter operates.
  5. Jun 3, 2004 #4
    Did you do a control (i.e null) experiment? For instance, have exactly the same set up except with cathodes which do not produce any thermal electrons at room temperature. Also, you could try putting in a barrier between the cathodes to catch the electrons.

    Also, I'm afraid that it is impossible for anyone to make any sense of your measurements as there are no indications of error bars. (you may think this pedantic but it is a very important issue)

  6. Jun 4, 2004 #5
    Reply to DrChinese

    Reply to DrChinese

    Thank you for your comments. I think the magnets are free.

    When I am going to do the experiment, I can borrow some magnets from one of my friends. After the experiment, I return the magnets to him and the magnets can be not changed at all. So, the magnets are free.

    In an ideal Carnot's cycle, a cylinder and a piston are used. Are the cylinder and the piston free? If we want to create a new cylinder and a new piston, we must spend some energy and so on. But it doesn't matter. You can borrow a cylinder and a piston from one of your friends before you start to carry out the cycle, and return them to him after you finished the experiment. They can keep unchanged totally. So we say the cylinder and the piston are free.

    Permanent Magnets are wdely used today in electric generators and electric moters, especially in those of powers range from several watts to hundreds of watts. These magnets can be used for one year, five years, ten years , twenty years, and even longer. Provided they are used carefully and not hurted accidently, they will be at last as fine as they were before they are used. There is no change at all in these magnets. So I think actually it is well accepted that magnets used in such ways are totally free.

    Would you agree my arguments?
    Last edited: Jun 8, 2004
  7. Jun 4, 2004 #6


    User Avatar
    Science Advisor
    Gold Member

    I am not certain. I agree it comes down to returning the magnet to your friend (in the analogy) in the same condition you received it.

    So let's ask: if you pull 8x10^-13A current at .08 volts out of the generator for 1.78 million years, you would get a kilowatt-hour. Would the magnet be in the same condition at that time? I think over that period it would be clear that total entropy obeys the 2nd law.

    But I do think the concept is ingenious even if a violation of the second law has not occurred.
    Last edited: Jun 4, 2004
  8. Jun 4, 2004 #7

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Doesn't this experiment effectively measure the energy needed to make the demon's decisions - by measuring the decreased field strength of the magnets over time, and dividing by the total number of electrons over that interval?

    My understanding of Maxwell's problem is that the paradox is resolved by the energy contained in structured information; as opposed to random data that contains no useful information hence no structure or order. The proof offered in one lecture was that a machine can be built that does useful work by reading structured information, but that can't do useful work by reading random data. From this, an information/energy equivalency is suggested.
    Last edited: Jun 4, 2004
  9. Jun 4, 2004 #8

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    The energy of the field in a magnet is restored as the object repelled or attracted reverses its motion. In the case of permanent magnet DC motors, the energy comes completely from the electricity supplied. No energy transfer from the magnets to the system is unidirectional. The system is always restored by means of momentum and electrical power losses. I believe that the back EMF is precisely this potential for work.

    The fact that we can't make a permanent magnet only motor is the proof. No free energy.
  10. Jun 5, 2004 #9
    I love this answer!
  11. Jun 5, 2004 #10
    Here are Maxwell's original words describing what later came to the called "Maxwell's Demon":

    James Clerk Maxwell
    Address:http://webserver.lemoyne.edu/faculty/giunta/demon.html [Broken]
    Last edited by a moderator: May 1, 2017
  12. Jun 7, 2004 #11

    The output current and voltage are really extremely low, but it means a new beginning, as an old restriction is broken here. Things can then be developed tremendously.

    In our present test tube FX1, there is only one pair of Ag-O-Cs cathodes. Actually we can settle several pairs of such cathodes in a single tube, and connect them in series to get greater electric motive force, or connect them in parallel to get greater current. If thousands or millions of such Ag-O-Cs pairs are settled on the surfaces of some quartz plates in a single tube, and connected in some proper way, a much greater power will be possible.

    Efforts should also be made to find new cathode materials with work functions lower than Ag-O-Cs. The work function of present Ag-O-Cs cathodes is about 0.7 to 0.9 eV. Can we find some new materials whose work functions are only 0.1 to 0.2 eV? There is no theoretical limitation to lower work function. The output power of a single tube may be possible to increase tremendously in this way.

    On the other hand, we can replace Ag-O-Cs cathodes to oxide cathodes (Ba and Ca oxide cathodes, etc), which work at a temperature of about 800oC. Such a tube can extract heat from a high temperature heat reservoir (a furnace or a nuclear reaction apparatus for example) and convert the heat into electric power, in large scales.

    When Faraday discovered electromagnetic induction in 1831, he could get only very weak and instantaneous current in his coils. He then made efforts of years to design or invent electric generators that could produce a strong and continuous current. What he succeeded finally was only something that can only be regarded today as poor child toys. Nevertheless, these poor toys are pioneers or pre-prototypes of the modern powerful generators that might posses a power of one million kilowatts each set.

    What we are doing now is trying to make “died energy” to revive again. Such a task is certainly much difficult than Faraday’s task of converting mechanical work into electric energy. There is nothing strange or surprising that the road in front of us is much longer than the road in front of Faraday.
    Last edited: Jun 8, 2004
  13. Jun 8, 2004 #12
  14. Jun 8, 2004 #13


    User Avatar
    Science Advisor
    Gold Member

    But you don't find magnets on that road, either. The magnet effectively powers the process, almost like a type of battery. Uranium contains an energy source too, but when you extract energy from it you don't violate the 2nd law. My point being that you assume that the magnet is not changed as work as extracted from it. I believe that the energy output (if any) is offset by matching changes in the magnet (if any).

    It takes energy to create a magnet (an ordered system). But there is no way to measure the strength of a magnet to a sufficient degree of precision to detect the subtle changes in the magnet. If you could, it would show that the magnet loses strength exactly equal to the energy output. Since the net gain is zero (or less than zero), the process obeys the 2nd law.

    That doesn't mean that the physical process you describe is worthless. I simply doubt that the 2nd law of thermodynamics is violated.
  15. Jun 8, 2004 #14
    The energy here is thermal. Not elecromagnetic.
    No, the magnet is not "depleted".
    It takes work, some expenditure of energy, to magnetize a magnet, but don't confuse a magnet with a battery, or a capacitor. The magnetic field is an inert tool that can be used to knock electrons around. It loses nothing in doing this. Magnets "wear out" because the little domains eventually slip out of alignment. No energy has been lost. A brief second in a strong magnetic field and it is good as new.

    Good magnets are those that have the most friction between the domains. These last longer. It is a matter of the material used, not the energy required to magnetize them.

    A magnet isn't "used up" by being employed an more than we are wearing gravity out by using it all day to stay stuck to the earth.

    The device described here is running off of thermal energy from the surroundings.
  16. Jun 9, 2004 #15


    User Avatar
    Science Advisor
    Gold Member

    1. I am not saying a magnet is a battery and my analogy in that regard was not a good one. What you are saying makes good sense. But I am not sure I fully agree with your comparison of the magnet and gravity in one sense. The magnet is an artificially ordered system and it takes energy to create it. It wouldn't surprise me if a way was found to utilize that order. Since this experimenter states that the current produced is some function of the magnet, it would seem a good place to examine when someone claims the 2nd law is being violated. (My confidence in the 2nd law is very high.)

    2. I am not sure what is creating the reported current in the described system, assuming there is one (which is probably a big if). Are you saying it is coming from the surrounding environment? I guess there could be a small temperature gradient in the apparatus which is not readily apparent (since the output is so low). If so, I wonder how that is converted to provide power.
  17. Jun 9, 2004 #16


    User Avatar
    Science Advisor

    In ferromagnets, the minimum energy is where neighbouring spins align. However, the maximum magnetic field energy of a macroscopic ferromagnet is at maximum field, when all spins are aligned. The competition between these two effects is what results in domains. So a soft ferromagnet like mild steel may have no magnetic field because there are many small domains whose fields cancel each other. In a "hard" ferromagnet, the spins do not change direction easily. So you can magnetize a chunk and it stays. (Good thing too, in the case of hard drives.) However, over time, even hard ferromagnets lose their field, because of the tendency toward randomization and minimum energy.

    I find DrChinese's suggestion interesting. Remember that the released electrons, in looping from one place to another, create their own magnetic field, which will act back on the permanent magnet and tend to demagnetize it.
  18. Jun 9, 2004 #17
    I understand your logic. Guided by Occam's Razor, though, we should approach all claims of violation of the "laws" of thermodynamics and motion as misunderstandings by the claimant. I have never encountered a situation where the law wasn't simply misunderstood, or where what was happening in a device hadn't just been misunderstood.

    The magnets here function analagously to a diode, taking a back and forth exchange of electrons between the electrodes and converting the motion to exclusive one way travel.

    The electrons, you will notice, he describes as "thermal" electrons. I take this to mean they are exited to jump off the electrodes by, or by something akin to, the photoelecric effect. Notice the special material of the electrodes silver, oygen, carbon. My guess is that this compound is more sensitive than most to having electrons exited off of it in this way. Silver puts me in mind of the fact that photo-sensitive silver nitrate is used in photography.

    However, the real evidence the effect is thermally driven come from his explanation of how the system tend to cool, and then to suck heat in from the environment to try to maintain a mean temperature. That's pretty explicit. That is where the energy is going into the system. It's not Maxwell's Demon. It's a solid state bobbing bird.
    Last edited: Jun 9, 2004
  19. Jun 9, 2004 #18
    This is the case to be sure. Using a magnet puts a strain on the domains, and encourages them to go out of alignment. I didn't bring this up before because I didn't want this effect to be misunderstood as "using up the magnets energy".
    How quickly a magnet's domains go out of alignment depends on how good the friction between the domains is. It is completely unrelated to how many joules it took to magnetize the magnet.

    The force of a magnetic field comes from the electric fields of the electrons you described. The electrons and their fields don't get "used up" or "discharged" in any way when you employ a magnet to do work. Nor is the amount of work you can do with it before it goes out of alignment related in any important way to the energy it took to align the domains.

    The newer (post WWII) ferrimagnetic magnets are just about immortal compared to the old hard-steel ones, so the wearing out of magnets is not a serious issue in all normal applications.
  20. Jun 9, 2004 #19


    User Avatar

    Staff: Mentor

    Sorry if this has been brought up already in this thread, but it seems like the original experiment is just a new version of a thermoelectric conversion. The magnet is helping to alter the paths of the thermal electrons so that there is a slight preference for the landing electrode. The effect could probably be maximized once the effect of the magnet was simulated, but the efficiency of the conversion from heat to electrical power probably would not exceed the traditional thermoelectric conversion efficiency, which is not considered high enough to violate the arguments against Maxwell's Demon.

    Some thermoelectric background if anyone is interested:

    Neat forums, BTW. -Mike-
    Last edited by a moderator: Apr 20, 2017
  21. Jun 9, 2004 #20
    Exactly, regarding the magnet.

    I am not certain what is prompting the electrons to fly off the electrodes in the first place, though. Their materal is very specific. I lean much more toward photoelectric as an explanation for these leaping leptons than thermoelectric.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook