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

Featured A Can indistinguishable particles obey Boltzmann statistics

  1. Feb 7, 2018 #1
    Many text books claim that particles that obey Boltzmann statistics have to be indistinguishable in order to ensure an extensive expression for entropy. However, a first principle derivation using combinatorics gives the Boltzmann only for distinguishable and the Bose Einstein distribution for indistinguishable particles (see Beiser, Atkins or my own text on Research Gate). Is there any direct evidence that indistinguishable particles can obey Boltzmann statistics?
     
  2. jcsd
  3. Feb 7, 2018 #2

    mfb

    User Avatar
    2017 Award

    Staff: Mentor

    The Boltzmann statistics is the high-temperature (or low-density) limit of both the Bose-Einstein and the Fermi-Dirac statistics.
    It is an approximation only, but a very good one in many cases.
     
  4. Feb 7, 2018 #3
    I forgot to say that I don't want to consider limiting cases such as high temperature. The combinatorics derivation I mention gives Boltzmann for distinguishable and Bose-Einstein for indistinguishable particles at any temperature. On the other hand, according to books, even particles obeying Boltzmann have to be indistinguishable in order to resolve the Gibbs paradox, independent of the temperature. There seems to be a contradiction, I would say. So my question is: Does anything else point to the possibility of indistinguishable particles obeying Boltzmann at any temperature and density?
     
  5. Feb 7, 2018 #4
    I have to add: For the model system treated in the mentioned derivation the Boltzmann distribution is an exact result, not an approximation. I don't know, however, whether there is a real system that is described sufficiently well by this model.
     
  6. Feb 7, 2018 #5

    mfb

    User Avatar
    2017 Award

    Staff: Mentor

    Indistinguishable particles cannot follow Boltzmann exactly at finite temperature. So what?
    Distinguishable particles are like many individual distributions with single particles summed, for these particles all three distributions are the same.
     
  7. Feb 7, 2018 #6
    According to many text books they can and do. Thus my question.
     
  8. Feb 7, 2018 #7
    I don't understand what you are saying here. Could you rephrase?
     
  9. Feb 7, 2018 #8

    mfb

    User Avatar
    2017 Award

    Staff: Mentor

    I'm sure you are missing the part where they call it an approximation.
    I don't understand what is unclear.
     
  10. Feb 7, 2018 #9
    I'm pretty sure I'm not. The argument in many/some textbooks is as follows: Distinguishable particles obeying Boltzmann give an entropy expression that is non-extensive and therefore wrong. If one includes a factor 1/N! to account for indistinguishability one arrives at the Sackur-Tetrode formula for entropy, which is extensive. Therefore particles obeying Boltzmann (such as atoms and molecules in a gas) have to be indistinguishable.
    If this was only true in limiting cases it wouldn't be very helpful since the Gibbs paradox wouldn't be resolved in the general case.
     
  11. Feb 7, 2018 #10
    The following sentence is unclear: "Distinguishable particles are like many individual distributions with single particles summed,..."
     
  12. Feb 7, 2018 #11

    ZapperZ

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    The Drude model for conduction electrons in a conductor is based on Boltzmann statistics. Isn't this a clear example of what you are looking for?

    Zz.
     
  13. Feb 7, 2018 #12
    Good point! However, in 1900 Drude had no idea that electrons were actually fermions and should obey Fermi-Dirac statistics. He simply described them as an ideal gas because that's the only thing he could do at the time. I'm not sure if that qualifies as evidence that indistinguishable particles can obey Boltzmann.
     
  14. Feb 7, 2018 #13

    ZapperZ

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    But what does history have anything to do with it? The Drude model is STILL being used, and in fact, it is the foundation on how we got Ohm's law! Many of the basic properties of conductors are based on such a model, i.e. single-particle, non-interacting model.

    In fact, in electron particle accelerators, we typically model charge particles using such classical statistics as well! Particle beam modeling packages do not use quantum statistics to get the beam dynamics.

    Zz.
     
  15. Feb 7, 2018 #14
    I haven't heard of this. Could you provide a reference?
     
  16. Feb 7, 2018 #15
    For example Blundell and Blundell: Concepts in thermal physics, 2nd edition, chapter 21, sections 21.3 to 21.5 and exercise 21.2.
     
  17. Feb 7, 2018 #16
    You have a very good point there. One could say that particles that are actually indistinguishable are described in a classical way (like distinguishable atoms in an ideal gas), and the results match experimental findings. That might be one answer to my question. Thanks.
     
  18. Feb 7, 2018 #17
    Reading this, I think I understand the confusion here. This book uses the terms distinguishable and indistinguishable more loosely than I assumed. It appears to assume distinguishable means something like, the particles have a different color or shape, and assumes indistinguishable means they are identical. The issue I have with this wording is that identical is not the same thing as indistinguishable. If I carefully watch an ensemble of classical identical particles, I can keep track of were each particle moved and what it is doing. Therefore, they are distinguishable. For quantum particles, I can not keep track of each particle as they interact with each other. Quantum particles are truly indistinguishable.

    When discussing Boltzmann statistics of a pure gas, the particles are assumed identical but distinguishable. It is really the fact that they are identical which leads to the need to prevent the over counting of states.
     
  19. Feb 7, 2018 #18

    Stephen Tashi

    User Avatar
    Science Advisor

    In another thread, @Andy Resnick mentioned a paper by Jaynes (e.g. http://www.damtp.cam.ac.uk/user/tong/statphys/jaynes.pdf ). Jaynes says that (Gibbs said that) whether to distinguish or not distinguish microstates is a choice made by the experimenter.
     
  20. Feb 7, 2018 #19

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    So I'm confused then. As mfb points out, Boltzmann is always an approximation/limiting case.
     
  21. Feb 8, 2018 #20
    I agree that identical particles can be distinguishable (in the sense of trackable). I actually thought Blundell saw it that way too.
    About your last sentence:
    I don't understand why the fact that particles are identical but distinguishable would necessitate a factor 1/N! in the partition function to avoid overcounting.
    You seem to be saying that swapping two particles in different states does not lead to a different microsctate even if it's obvious that the particles have been swapped. My understanding was that swapping distinguishable particles in different states leads to a new micrstate even if the particles are identical.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted