Question:
Statistical mechanics: is it too daring to say this about freezing-out of d.o.f.?
☮ Vašek
2010-05-07 06:56:16 UTC
I am writing some notes for second year university students, in which I discuss the topic of an ideal gas with internal degrees of freedom (i.e. taking into account the rotation and/or vibration of molecules). They learn that for temperatures lower than some characteristic ones, the classical description is no longer valid as this kind of motion "freezes out" due to quantization.

Is it possible to say that when the translational degrees of freedom freeze out, we are facing Bose-Einstein condensation (for bosons) or Fermi degeneration (for fermions)? Or is it too daring a comparation? I think that the idea is exactly the same: that the quantization of phase space (in vague terms) starts to manifest at extremely low temperatures, not allowing us to talk about continuous position & momentum any more.

On the other hand, I don't want to feel responsible if they come up with this at an exam and are told that it is nonsense.

Do you think I could add this claim in the notes?
Four answers:
?
2010-05-07 13:30:18 UTC
Rotational and vibrational degrees of freedom are, so to say, 'Boltzmann factored away' because of the finite energy difference between ground state and excited states becomes so large wrt to kT that the exponentials are 0 for all practical purposes. You can't say the same for continuous degrees of freedom, or you have to adopt very speculative views about the structure of space on small scales. So I'm afraid this kind of comparison for beginning students might be risky and could

induce misleading ideas. Moreover, one can't speak of freezing out of translational d.o.f. You can think of a nucleus as a Fermi gas at 0 temperature, but Fermi motion is quite effective as can be checked, for example, by looking at the momenta distribution of a beam of particles scattered by these nuclei..



Edited 1: one should not overstress, either, a link between undistinguishability, quantum behaviour and low temperatures. It is true that undistinguishability manifests itself in special ways in the quantum realm because of the symmetry/antisymmetry of the wave function of a system of several identical particles, but a simple hydrogen atom is a totally non classical system and its quantum behaviour is not linked to any undistinguishability. On the other hand, the cosmic maser example recalled by FGR reminds us that low temperature is not required to observe coherent quantum effects. (but of course it helps in condensed matter..)

The random walk in the complex plane evoked by Scythian takes us back to what is (or so I think) the most enlightening novelty brought by Feynman to quantum physics, namely the sum over paths. It is not always practically useful in ordinary qm (it is mandatory if qft) but it does bring with it a picture which allows to show that indeed there is a kind of continuum between quantum and classical behaviour. You can always calculate a transition amplitude by summing over paths. And in the classical case (noise provoking destructive interferences or blocking of most paths) it will reduce to a sum over neighbouring paths and finally to a delta function in terms of the probability distribution of the final states, i.e. to the single classical solution. But the continuity of physics is evident here, which it is not at all if you approach qm the operator way.



Edited 2: Scythian, I think you are mixing two different categories concerning qm. One is mathematical formulation, the other is interpretation. Mathematical formulations are the backbone of the theory and although there exist quite a number of them, they are all equivalent. Feynman's sum over histories is a mathematical formulation. Not always the most manageable one and sometimes unduly cumbersome for elementary problems. On the other hand, it is mandatory in gauge field theories where the operator formalism leads to intractable calculations. Still, wherever several have been used, all mathematical formulations lead exactly to the same results, as did, from the very beginning, the matrix formalism of Heisenberg and the wave formalism of Schrodinger.

On the other hand, people unsatisfied with a positivist view point which considers that there are mathematical schemes and observations and that all a physical theory must do is to build mathematical schemes which account for these observations, have spent much time building interpretations. Everett's branching worlds pertains to this second category. It provides no new mathematical scheme and predicts nothing which is not already predicted without it (and this is true of all 'interpretations' of qm) Its main point is to shun the problem of the reduction of the state vector by positing that there is no such reduction but an infinite branching process, which encompasses the observer himself so that he keeps having the impression that the wave function has indeed collapsed (his many branches being unaware one of the other) but at the same time the more 'enlightened' physicist 'knows' that there is no collapse and that all possibilities are simultaneously realised.. . it's not ugly but from a practical view point it brings strictly nothing.

My viewpoint is that, although it is a mathematical formulation Feynman's sum over histories is far more intuitive than operator formalisms and that it gives you the feeling to have a grasp on 'reality' at least as far as single particle problems are concerned. You don't need to wonder about which way or about wave or particle in the double slit experiment to take the simplest example.

The sum over paths allows ( me, maybe its personal..) to build a truly satisfactory picture.



About indistinguishably: the Asker explicitely considered degeneracies concerning assemblies of identical particles and that's what is usually understood by indistinguishability in qm. It is only for identical particles that the phenomena he wrote about take place.(Bose Einstein condensation, Fermi Dirac degeneracy -and I don't see, as already said, where a freezing of translational dof comes in here)

I don't think it very useful to extend the meaning of the word undistinguishable. The risk is to introduce semantic ambiguities and arguments where none exists or is necessary.



Edited 3: I just noticed that you (Scythian) had made explicit the distinction between formulations and interpretations in your comment. The fact is that I read it at noon and answered without going through it again eight hours later. So, sorry for lecturing. BUT the main point, that Feynman's sum over histories is a formulation and therefore not akin to Everett's interpretation remains, and so does my contention that the picture brought by Feynman is far more clear than the picture -if any- brought by operator formalisms.

And I don't understand your strange slip from undistinguishability to uncertainty in position -which is a completely different concept- to the many world interpretation. By the way, string theory is not (contrary to quantum field theory) an achieved theory (not sure achieved is the right word) I'm not even talking about predictive power or relevance to the actual world. It is perfectly allowed and useful to play with model theories which are not the choice of Nature as training grounds, theoretical laboratories and the like. But string theories still lacks some basic pieces to make them complete logical and mathematical constructs.



Edit4: Vasek, the remark you quote (no dof freezing there) was alluding to the second case (F-D) and more precisely to the observation I made earlier about the substantial Fermi momenta of nucleons in nuclei. As for B-E, the point is that below Tc you only have a part of the particles in the p=0 state since you can't reach T=0. Of course it can be a sizeable part but that was my starting point: because of the discrete energy spectra of rotational and vibrational excitations, you can cool them away completely . Ok, the exponential are not mathematically 0, but still, that's far more 0-like than (T/Tc)^(3/2). Another point is that there still is a collective translational dof, but nothing like that exists for vibrations or rotations. Still, your cas is certainly better with bosons than with fermions.



Scythian, I would love to discuss entanglement but maybe we should set up another question for that.

My viewpoint has to do with the fact that, apart from non interacting gases, systems have properties that do not stem from the sole consideration of their components. We can study the system or its components, but not both completely and simultaneously because we destroy the system when looking at components. This translates into qm by the impossibility of speaking about the spin projection of a spin 1/2 particle in an EPR like S=0 state. If you insist on measuring it, you destroy the state without revealing any of its properties. You cannot get anything else than +/-1/2 but what you get has nothing to do with the state you have destroyed. Therefore this 'measurement' is by definition devoid of any information content. (maybe qm is the only place where you can make such meaningless measurements. We should explore that) . Of course you instantly know the result of a potential -but meaningless- measurement done at a space like separated event. But given that they are both non-events, it should not be too chocking. Asher Peres once wrote that "unperformed experiments have no results " ( ;-) in fact about the same kind of EPR situation where, if you assume that you can speak meaningfully about what the results would be, should your device be positioned at a different angle, you derive Bell's inequalities and contradiction with Nature.

So maybe the very fact that the measurement of the far electron of the pair does have a result even if unperformed (once you have made the measurement on your side) tells us that it is a non experiment. ( It might also be that I'm simply tired. I'll stop here for now.)
Frst Grade Rocks! Ω
2010-05-07 14:17:34 UTC
Nice dialectic between you and Scythian. Here is some more food for thought (I will diverge a bit from Scythian and you initial discussion).



I think it is a great idea to state the limits of classical analysis and to point the direction one must head in order to understand what is going on. Don't worry too much, they are all aware that everything condenses so classic ideal gas law doesn't work in cold environments



This is the realm where quantum mechanics begins to have a noticeable effect on vibration and rotation of molecules. Gas molecules have fixed states of vibration and rotation. And they emit fixed frequency electromagnetic radiation in dropping back down to their ground state just like atoms do when electrons change states.



There are several cutting edge examples of this in physics. One is studying the maser effect of cold molecular gas clouds in space. If the gas molecules are stimulated to a higher rotational or vibrational level, they fall back to the lower level and emit microwave radiation. And they do so in unison creating a microwave laser. http://en.wikipedia.org/wiki/Astrophysical_maser



*I like masers because that is an area that directly melds classic ideal gas with qm of vibration and rotation. BEC might be a little "cooler" area of study, but it is also a little farther removed from ideal gas your students are studying. (Then again, Einstein came up with the idea of masers as part of his work on BEC)



Another example is the study of BEC where again molecular vibration and rotation is in its lowest quantum level and the molecules act as a coordinated body rather than a set of independent molecules. (hate the idea of using the word "freezing" -- lowest quantum state works just fine)



It also deals with the greenhouse effect. CO2 has an absorption and emission spectrum in the infrared. This spectrum is governed by the quantum states of rotation and vibration. (I'm not sure if we are dealing with the ground state of CO2, but regardless we are dealing with quantum states).



Applying quantum mechanics to vibration and rotation of molecules has interesting results.
seamons
2016-10-04 13:50:51 UTC
Um that is beautiful flawed. You could wish to thaw out the chook first and prevent from a few embarrassment if you're going to do it in entrance of persons. And sure I could say you ordinarily must put on a condom. You would turn out to be with a few nasty infections. Have amusing.
Scythian1950
2010-05-07 08:50:24 UTC
There is a strong connection between freezing and emergence of quantum properties. Quantum behavior occurs once indistinguishability becomes dominant, which also characterizes Bose-Einstein condensation or Fermi degeneration. In a "hot, noisy" world, indistinguishability breaks down and decoherence takes over---every particle has its own set of attributes easily resolvable from those of others. However, freezing a sample is the same as producing a "quiet" quantum state so necessary to produce quantum behavior, such as quantum computation, which is why much of early quantum computation experiments have been conducted in ultracold conditions. I don't think it's a stretch to say that classical description fail at very low temperatures, when such classical motion (based on infinite degrees of freedom) is "frozen out" due to quantization.



Quantum theory in fact holds that even you or I are actually in quantum superposition, i.e., we exist "smeared" over time and space. But this range of smear is so tiny as to be irrelevant and undetectable under ordinary circumstances. However, if we were frozen and left alone in a vast, cold, empty space (think of this universe extremely far into the future, hundreds of trillions of years from now, if it turns out that it continues to expand), we would actually become like solitary, undisturbed electrons that exist in a quite broad volume of space. It would not longer be true to say that we'd exist in one particular place and time any more.



Edit: After some thought about the way you worded your question, I think it should be noted that the concept of quantization is something too easily misunderstood by many not familiar with quantum theory. Quantization is strongly related to the problem of indistinguishability, not an "inherent property of time and space", like somehow at some very small dimensions space and time is some kind of a checkerboard. It would be misleading to argue that as things get cold, everything wants to "settle into some kind of a lattice-work". We know that's what crystals are, but crystals do not form because of "quantization". We need to get away from that kind of highly misleading interpretation of quantization. I think a better analogy, as a starter, is to imagine a stream of water from a faucet, as it is slowly reduced. The stream is steady, and grows thinner, until finally it can only come out in distinct drops---and that is the point of quantization. The water comes out in countable drops.



Edit 2: Per your added comments, can you restate the question more precisely? I was looking at this from a pretty general point of view. By "second year university students", I assumed you were talking about introductory quantum mechanics.



Edit 3: If this is indeed an introduction to quantum theory, I would stress that physical systems are typically "in a classical state", or "in a quantum state", each of which exhibit different behaviors requiring different mathematical apparatus to treat. It is possible to "bridge" from one to another, as with the example of Rydburg atoms, where the outside electron orbital is so excited that the electron actually starts to resemble a classical satellite orbiting the nucleus. In other words, it's not necessarily an "either-or" between the two states, classical or quantum. But the progression towards quantum strongly involves the question of indistinguishability, and freezing is one way to create conditions for such indistinguishability, just not the only way. As another example, Feynman had suggested that quantum behavior can be mimicked by a "random walk on the complex plane", with very good agreement with the solutions of the Schrodinger wave equation. The interesting thing is, when "noise" is introduced to this kind of random walk, by slightly increasing amounts, this random walk begins to resemble that of the classical random walk. In fact, it should be pointed out that the only difference between the Schrodinger wave equation and the heat diffusion equation is the imaginary number i. Something to think about.



Edit 4: I liked the answers given by jean and FGR, so I've given both of them TUs. Hopefully, all of this will help you to decide how to handle this in the matter of your notes to 2nd year students. However, it's been in my experience that people new to quantum theory (and even some of those already familiar with it) have misguided notions of quantization, thinking that space time has some kind of an inherent graininess to it.



Edit 5: A nice PDF paper on the subject of Hadamard Walks is given in the link below. This is a more direct approach to quantum random walks, which does allow "bridging" to classical behavior. This is an effort at a better conceptual understanding of the link between quantum and classical behavior.



Edit 6: In continuing on this interesting subject, I'd like to respond to jean's statement that a single hydrogen atom exhibits quantum behavior not because of "indistinguishability". But in fact, that's a timid way to look at it, limiting the concept of indistinguishability to only indistinguishable particles, leading to the phenomenon already noted by the Asker. Indistinguishability can be argued even for a single particle, when it comes to resolving its actual location at a specific time. Another way of looking at this is Everett's Many Worlds hypothesis, a popular one that imagines that, say, a solitary particle exists in many parallel worlds, each with its own history, which is another way to say that we're dealing with many virtual indistinguishable particles that, together, behaves as a single particle. Feynman's sum--of-histories is analogous to this model as well. Quantum theory, while still poorly understood from a conceptual point of view, does allow a number of mathematically equivalent ways to describe it, all delivering the same results, even if the "explanation" are different with each one. This is the reason why "quantum interpretations", which attempts to develop an intuitively understandable concept of quantum behavior, is a field of study separate from main quantum theory, since main quantum theory exists on the strength of the mathematical apparatus involved, and doesn't actually require an "intuitive explanation" for it. And that is part of the problem with the question as posed by the Asker, because what the Asker is proposing to do is to offer some kind of an "intuitive explanation" for emergence of quantum behavior.



I think one of the hardest things for people to understand is how it is possible for different models with seemingly conflicting "explanations" can actually all be in mathematical agreement. We naturally assume that there can only be one "correct version of the events", all the others wrong, as for example, we don't accept the idea that there can be more than one way "the world was created", which was actually not uncommon in ancient societies. Yet, this commonsense rejection of multiple explanations for the same thing doesn't apply in physics when the mathematical apparatus can be demonstrated to be equivalent. I personally don't like Everett's Many Worlds hypothesis, or even the String Theory model (where everything is made up of tiny vibrating strings), but that does not mean that I believe that the mathematics involved is in error and so those models must be rejected. I just go over to alternate ways of looking at the same things without invoking those conceptual models that I find disagreeable. I think for many people, this is a notion hard to swallow---actually having a choice of how you'd like to think of physics conceptually.



Edit 7: jean, I'm just expanding the subject, I know I'm probably far beyond Vasek's original question, and into highly speculative issues. But everything that I have brought up is already in the literature. For Vasek's original question as narrowly defined, I'm deferring to your answer. As for your objections to "confusing" indistinguishability with quantum uncertainity, again, this is not my original or novel idea, but it does fall under quantum interpretations. The main problem with quantum interpretations is that almost by definition they cannot be experimentally verified, because it does not seek new mathematical formulations of quantum theory, it merely attempts at a conceptual or intuitive understanding of it. Again, I think it's a too-timid approach arguing that indistinguishibility and quantum uncertainity are "two entirely distinct matters".



Do you have any idea how to conceptually explain quantum behavior? How, for example, does quantum entanglement even work at all, conceptually? We know it works per math, but can one make any conceptual sense out of it?



Edit 8: Vasek, that's interesting that quantum walks has been your field of expertise for some years. I'm sure you're aware that even with this one subject, there are several different branches or avenues of study, and that your particular expertise is associated with QFT, since that's what puts the bread on the table.



As for your comment, jean, on string theory "not an achieved theory", ironically, the more a theory is mathematically consistent with other established theories, the harder it is to demonstrate experimentally the distinction it has! String Theory is still a remarkable mathematical model in its own right, and as such, it shouldn't be discounted, but at least seen as an "alternate way of looking at things". Recall the early original "conflict" between Heisenburg's Matrix Mechanic


This content was originally posted on Y! Answers, a Q&A website that shut down in 2021.
Loading...