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

What does a BEC look like?

  1. Jan 13, 2009 #1
    What exactly does a Bose-Einstein Condensate look like with the naked eye? Is there any special about what it looks like through the optical part of the spectrum? What are the electromagnetic properties of a BEC as a whole?
  2. jcsd
  3. Jan 17, 2009 #2
    I never saw one, but this is from what I understand:

    You start with a diluted gas. Most gases are transparent. Chlorine for example isn't, but if you see it, you notice, that the particle density is so low, that it is still mostly transparent. Then you cool it a lot, you have it in vacuum and evaporate a lot. Finally you need to put it into the focus of six lasers with high intensity. For cooling below a few tens of Kelvin. You don't want random light in your apparatus as it would heat things.

    So I would guess: You are dealing with a tiny amount of atoms, that is too small to be visible. If it was visible, you wouldn't want to look at it because you would introduce outside light, and that could destroy the condensate. If you could look at it you had six powerful laserbeams focussing that would overshine anything you could see.

    But maybe I am wrong and maybe I remember the question when I meet a friend who deals with that stuff...
  4. Jan 17, 2009 #3
    I met the friend today. The condensate is large enough to see with a camera. He didn't know what the magnification was. But it was just a smeared homogenous blob glowing in the laser light.
  5. Jan 18, 2009 #4
    At a first guess, the condensate will behave like a dilute atomic gas. Basically, the visual appearance is strongly controlled by the electronic structure, and I don't think condensation changes that --- after all, if molecular formation occurs, the molecules would leak out of the trap. The trapping procedure itself tends to cause radiation (optical molasses basically relies on bouncing photons off the atoms), so at that frequency, it would be possible to see it as a glowing blob. It should also be possible to see it distinguished against the background in absorption, i.e. backlit. I seem to recall seeing pictures of vortices in BECs, where the vortex line is visible because it's basically a complete lack of particles in the BEC.
  6. Jan 22, 2009 #5
    Someone asked somewhere about the density of BECs Ketterle gives a figure of magnitude at some MIT page at 10^14 particles per ccm.
  7. Jan 22, 2009 #6
    This is a little like a zen koan. Is what you see still BEC?
  8. Jan 23, 2009 #7


    User Avatar
    Science Advisor

    Then have a look at a microcavity polariton BEC. Here the emitted light is a direct part of the BEC.
  9. Jan 23, 2009 #8
    Really? So the particular atoms that emit the light remain part of the BEC phase afterward (rather than decohering out)?
  10. Jan 23, 2009 #9


    User Avatar
    Science Advisor

    The particles you want to condense in microcavities are the polaritons, which are the semiconductor equivalent of two coupled classical oscillators. You get all the same stuff like in the classical case like mode splitting and so on.

    Now the two modes, which couple in a microcavity are the photons and the excitons, so you get a bosonic quasiparticle, which is a bit photon-like and a bit exciton-like. In the interesting case in the middle of the k-dependent dispersion the polaritons are about 50% each. If you get these polaritons to condense, the escaping photons are therefore directly reflecting the characteristics of the bec as there is no conversion process involved. Some of the polaritons just leave the cavity and you register them. Due to the small mass of these polaritons BEC is expected to show up even at room temperature. The drawback is, that the lifetime of these particles goes down as the photon fraction increases. The average lifetime can be as low as a few ps in a condensate, so the BEC will be a nonequilibrium condensate as you need to "refill" the escaping polaritons.

    Additionally I should add, that although the publications of last year are rather sure that polariton BECs are real BECs due to seeing quantized vortices, Bogoljubov like excitations and other stuff showing up in usual BECs, I am still not really convinced. For example there was still no conclusive and quantitative measurement on the intensity correlation, which would clearly demonstrate a spontaneous phase transition.
  11. Jan 23, 2009 #10
    hadn't heard of polaritons; interesting.
  12. Jan 23, 2009 #11
    I read that if the condensate is placed in e.g. bowl, then some of the condensate will actually try and escape outside the bowl, because the potential is lower outside the bowl. So the BEC will look like a sine wave just as we know it from the potential well in quantum mechanics,
  13. Jan 23, 2009 #12
    If you mean a literal bowl the condensate shouldn't survive for long. Making solids that cold is impossible afaik.
  14. Jan 23, 2009 #13
    Unfortunately I can't remember where I read it since it is a long time ago, but I am quite sure they meant it in a hypothetical way i.e. just to emphasize that we are dealing with a macroscopic object that can be described by Schrödingers equation.
  15. Jan 23, 2009 #14
    Cthugha; I haven't kept up with solid state bec's ....but..

    Would this mean that there is no BEC in the cavity or that it simply is not lasing.??
  16. Jan 23, 2009 #15


    User Avatar
    Science Advisor

    I don't know, whether I got your question right. If I completely misunderstood you, please just ask again.

    Well, unambiguously showing that a spontaneous phase transition is happening would be a rather good indicator that you really have a BEC. What one would need to show is, that with increasing excitation density the photon statistics changes from thermal behaviour (indicated by "bunching" g2 (tau=0)=2) towards some coherent state (indicated by g2(tau=0)=1) without destroying the strong coupling regime (because then you just have an ordinary cavity laser). However bunching will only show up on timescales on the order of the coherence time, which is pretty low in most semiconductors, maybe 20 ps. In fact the temporal resolution of usual photo diodes is not good enough to reflect the real photon statistics. There have been measurements of g2, but due to the bad temporal resolution one just sees values of around 1.06 instead of 2. So these experiments do not really give quantitative answers. I think, this is just a matter of time, but who knows for sure.
  17. Jan 23, 2009 #16
    I'm not sure I asked it right; like I said....I haven't kept up with solid state.

    Ok, so condensation is equivalent to polariton lasing, (but similar to 'single pass' lasing), no?

    So empirically, if the detected photon is characterictic of the polariton BEC, and is distinquishable from photonic lasing, why the ambiguity ? Is bunching the only signature? or is there a polarization condition?

  18. Jan 23, 2009 #17
    Liquid helium will leak over the rim of a bowl...
  19. Jan 24, 2009 #18


    User Avatar
    Science Advisor

    Yes, that is the major opinion and I think so, too. Polariton lasing is supposed to work due to bosonic final state stimulation and is equal to condensation.

    Well, the main problem is the question, where the coherence comes from. There ae three main ways to get the condensate: nonresonant excitation, resonant excitation at k=0 or stimulated scattering by resonant excitation near the magic angle, which starts an OPO-like process.

    The first method is a bit complicated as the polaritons need to relax down to the ground state by polariton polariton scattering. As the scattering processes are rather slow compared to the polariton lifetime and there is a relaxation bottleneck at high k values of the lower polariton one needs quite high polariton densities to get enough scattering to have a sufficient population in the ground state. In most interesting materials (GaAs) the polariton density needed for this process is already so high that the strong coupling is broken, polaritons are not bosons anymore and you will just see usual photon lasing. It is supposed to work this way in CdTe and some other materials with large oscillator strengths and in these materials some interesting results have been found like high populations inside the ground state and rather large spatial coherence.
    The other two mechanisms are more commonly used, especially the OPO configuration. The drawback here is that stimulated scattering and direct resonant excitation are coherent processes. So you see a coherent population build up in the ground state, but you never know, whether this property was just taken from the coherent pump beam. So these are good for oractical purposes, but not really good for prrof of principle studies. Although a clever trick came from the Snoke group. they applied mechanical stress near the position of the pump beam, so that you have a potential trap there. Afterwards one sees that the population migrates from the pump beam towards this spot, so one can guess that the coherence from the pump laser should be lost. However there are still critical voices around, mainly from the guys doing conventional VCSEL theory, that all of these features are not necessarily signs of BEC, but could be explained by some other kind of lasing process happening. Second order intensity correlation measurements under nonresonant excitation conditions however would be a rather unambiguos proof of the BEC.
  20. Jan 25, 2009 #19
    Thanks for your informative response. I was unaware of the various pathways to polariton condensation.

    Related questions ...

    1. what is the polariton spin state at k=0.

    2. Pragamatically; Have there been any temperature measurements to empirically establish a critical temperature (of phase transition) above which apparent condensation effects disappear...{since low effective polariton mass ( in the range of photon eff. mass) seems to make it canidiate for high temp. condensate of high efficiency?}

  21. Jan 25, 2009 #20


    User Avatar
    Science Advisor

    If I remember correctly, there is also a spontaneous buildup of polarization at the condensation threshold. I suppose this happens due to bosonic final state stimulation. Once one of the two spin states has a higher population, scattering rates into this state will rise as well.

    This depends a bit on the material system you are looking at. As soon as the thermal broadening exceeds the Rabi splitting, the strong coupling and the BEC are gone. There have been claims of polariton lasing at room temperature in bulk GaN microcavities, but I am not sure, whether the authors of the papers could demonstrate that they were still in the strong coupling regime.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook