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

Questions on Thin film interferrence (split from Snells law thread)

  1. Aug 16, 2003 #1
    There are two things that bother me about the reemission idea.
    The first is that the reemission has to be in discrete frequencies. So you'd think that many atoms together cause degeneration so that essential any frequency (say in the visible range) can be emitted.

    Ok so far but what about thin films? There might be only a few atom layers thick. Can you still get essentially any frequency out that came in?

    Second thing is wouldn't some photons get lucky and avoid hitting any atoms at all? Then there would be no slowing down at all and you would get 2 speeds for photons, one fast and one slow. Has this been observed?

    If this belongs in another thread feel free to move or ask me to start one.

    I will buy QED but I won't be able to read it until Dec so I hope you can address these issues now.
    Last edited: Aug 16, 2003
  2. jcsd
  3. Aug 16, 2003 #2


    User Avatar

    Staff: Mentor

    Yes, yes, yes, yes, and yes.

    Ever see a picture of an astronaut's visor? They are coated with gold. The coating is thin so some gets reflected and some gets transmitted.
  4. Aug 16, 2003 #3
    Reflection and transmission off of a thin film would happen whether or not the photons just pass on through or are absorbed and reemitted. So this doesn't help much.

    Since you said yes, that the phenomenon of some photons passing through at speed c and the reemitted ones are delayed has been observed can you give us an example please?

    It seems this would cause a great deal of distortion to a rapid pulse of light traveling through a transparent medium. Maybe you could even turn one pulse into two.
  5. Aug 16, 2003 #4

    I believe that is what Integral and I was aluding to with the Feynman "amplitude" discussion. I was put back because Feynman didn't go any further than talking about amplitudes, i.e., he didn't talk about individual photon/electron interactions where the photon absorbs and re-emits a photon, but noted that it might well be that he couldn't.

    I think what you actually have to do is consider that the photons react to the EM fields of the electrons in the sample, and are scattered about because of that rather than discrete absorbtions and re-emissions. In that way all the photons would act the same.
  6. Aug 16, 2003 #5

    Another thing. The thickness of a sample by itself it not always telling .. Some of our best reflectors are very thin.

    I've never seen any discussion or experiments along this line .. I'd like to to see one, cause I'd almost bet that in every instance of light going through a medium, you're going to wind up with at least some of each of these:

    Some of the light will be reflected,
    Some of the light will be refracted,
    Some of the light will be absorbed,
    Some of the light will be diffracted,
    Some of the light will go through unaltered.

    And it's all just a matter of degree as to what comes up on top. That is, that you'd have some of each, but some of those might be miniscule.
  7. Aug 16, 2003 #6
    You are exactly right about the first 3 but I'm not sure what you mean by 'go through unaltered'. Diffraction is a different phenomenon so let's put that aside for now.

    The best descriptions of these process where in my fields and waves classes ( one text is Electromagnetic waves by Umran Inan) which included lossy media, and in Optics (one text is Intro to Optics by Frank Pedrotti) which had more thin film examples but no lossy media.

    Both these books are senior or 1st year graduate level material so they are not easy reading.
  8. Aug 16, 2003 #7

    Well, by "unaltered", I guess I mean "none of the above"! ;)
    No, I just mean that it goes on the course that it was going on before getting into the new medium, uneffected by the medium.

    Also, I don't see where diffraction could be excluded. It is simply a spreading out of a light wave. Though, most texts, to demonstrate its effects, show diffraction as a pattern of light bending around corners, and/or in conjunction with interference. It kinda tends to make you (me) believe that is the only place where it does its stuff. Light has to have some interaction with a medium to diffract; I don't know of any reason why some of that diffraction can't occur inside the medium.

    Now, that's just my take on it, that all of those would happen when light enters a medium. I can't point to a reference that nails it.

    You know, in the QED book by Feynman, there is a part on partial reflection of light going through glass. Depending on how many layers of glass there are and how thick the layers are, different amounts of light will go through the glass. In one configuration he talks about, he says no light will go through, all of it will be reflected. I don't know what to make of that, in regards to what I said that all of those 5 things probably happen to some degree. Its hard to tell -- he may not have been talking in absolute terms that absolutely no light got through; only that it was virtually no light got through and virtually all of it was reflected.

    Yep, I know what you mean about the "not easy reading". I've never had a physics course in my life. The little I do know about it has come about by reading, re-reading, and more reading, over the 25 years or so since I got through with schooling. The Internet has been great help in understanding some of the things I have read but don't quite understand. Physics is such that you don't meet too many people in your area that like to discuss it! But the Internet is able to bring in people from a very wide area together.
  9. Aug 16, 2003 #8


    User Avatar

    Staff: Mentor

    I'm not sure there is a good example. Since the film is thin, the delay is very short. But the fact that you get both reflection and transmission says the same thing.

    edit: wait, I just remembered a GREAT example. The sky. The color of the sky and the sun in the sky depends on how much of the atmosphere the sun is shining through (its angle). Due to the thin-ness of the atmosphere, there actually isn't much scattering or transmission and how much of the atmosphere the light goes through makes a big difference.
    Last edited: Aug 16, 2003
  10. Aug 16, 2003 #9

    I don't think that refraction (or reflection) can be accurately modeled as electrons absorbing and re-emitting the photons.

    Consider this:

    Once the electron has absorbed the photon, it has a very high amplitiude for emitting a photon, but it is my understanding the timing of emission, and thus the direction of emission, is totally random. There are a couple of things there:

    1) in refraction there is a preferred (definate) axis the light transverses the medium,

    (2) if there is random emission, then the photon might get thrown back up into the medium, or sideways into the medium. It could have another interaction and get scattered even more back up into the medium. If is was lucky to get back into that refraction axis, it would be delayed from other photons that didn't get scattered about so.

    (3) depending on the thickness of the medium, you could expect multiple absorbtions/re-emissions per photon. Each one would slow the effective speed of the photon down. If absorbtion/re-emission is to account for the slowing of the light, then the light should be slowed even more depending on the thickness of the medium.

    So, take light and shine it through a medium, but pulse it with a very short pulse. The light should be coming out of the medium over a longer period of time than the pulse, as some of the photons got delayed more than others.
  11. Aug 16, 2003 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Don't forget that there are billions (trillions?) of chances for a photon to interact with atoms in a medium. We can, I imagine, appeal to the law of large numbers, and state that the most likely photon behavior is virtually certain to occur.

    Though, I suspect that there is some nifty way to apply conservation of momentum to state a stronger conclusion.
  12. Aug 17, 2003 #11
    As blue light is refracted less than red light, does it mean that blue encounters less atoms on its path and is thus less delayed?

    wait, isn't blue sky due to scattering from nitrogen, but red sky due to differing refraction angles for different colors?
  13. Aug 17, 2003 #12


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    The "most likely outcome" is different for red and blue light; red light is most likely to be transmitted while blue light is most likely to be scattered.
  14. Aug 17, 2003 #13
    For me, the issue isn't the existance of reflection and refraction, that's obvious. The question is are you really saying that when passing thru a dielectric medium, some photons are 'delayed' by absorbtion and reemission and some travel at c because they happen to never hit an molecule?

    If you say yes, than there are 2 speeds for the photons in the medium. I have never heard of such a thing. Can it be demonstrated in a lab? Your sky example doesn't answer the question since we can not tell how long the photons are taking to reach us. We need controlled conditions where we can time their arrival and see photons that continue to travel at speed c through plastic, etc.
  15. Aug 17, 2003 #14


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    It would be the rare photon that makes it across a surface uneffected. Unless of course you are talking about a vacuum.
  16. Aug 17, 2003 #15
    Just when you thought you heard it all, here's another description.

    When a photon interacts with a transparent medium it interacts coherently, that is with many molecules at a time. You can think of it as "sharing" some of its momentum with the medium, so that it temporarily becomes a massive particle. In an optical fiber its mass prevents it from tunneling out of the fiber and being lost.

    If it has more momentum (shorter wavelength) it needs to share less with the medium, and that means it displays dispersion, shorter waves travel faster.

    So it isn't correct to think that it interacts molecule by molecule, but layer by layer of molecules as it moves through.
  17. Aug 17, 2003 #16
    Your description reminds me of the effective mass of an electron in a conductor.

    If a photon interacts with a few molecules per layer than it's like having a large cross section and it is very unlikely not be slowed down even by a single atomic layer. Therefore the relative velocity would not have any dependence on material thickness as expected.

    This description also agrees with the simpler approach of stating the relative velocity and not worrying what the photons are up to but assuming they are all slowed down. I like it. :smile:

    Can you cite a reference Tyger? You've peaked my curiosity.
  18. Aug 17, 2003 #17


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I knew that effect could happen (interacting with many molecules), I didn't know if the "normal" situation was well approximated by single photon - single electron interactions. (Can you expound upon this, Tyger?)

    I came across the effect myself (or at least something similar to it) when trying to find out what was going on in the (fairly) recent experiments about scientists transmitting a light wave "faster than light" (it was, IIRC, the phase velocity, not the actual velocity of a photon), and the experiments where they greatly slowed a light wave, even stopping it entirely. The best description I could find was here
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook