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

Photon collision

  1. Jan 2, 2007 #1
    whilst we are successful in colliding matter to identify yet further fundamental particles, and being aware there is a limit on speed due to mass end energy problems, and given that matter and mass are interchangeable, might it be possible to collide a single photon with another (this must rely on the particle nature of light). How could this be done? Ultimately every fundamental particle has a wavefunction and asscoiated energy. So if we remove the problems of mass can collision of photons show any new discoveries? Oh, and please do not go on about c being infinite and independant of direction etc, the question simply targets collision of photons.
  2. jcsd
  3. Jan 2, 2007 #2


    User Avatar

    Staff: Mentor

  4. Jan 2, 2007 #3


    User Avatar
    Science Advisor

    What would a single photon scatter?
  5. Apr 26, 2009 #4
    How about creatng a virtual vacuum using two plates at distance < wavelength of visible light, using a full spectrum light source, a single slit filter just large enough for a single photon to pass through at a time, and using Casimir effect, measure the electrical potential difference between the two plates. If a photon photon collision occurs, an electron positron pair will result causing an electrical potential difference between the two plates, the potential difference will cause an oscillating pulse, indicating the energy of the photons.
  6. Apr 26, 2009 #5
    Photon-photon scattering makes sense and its cross section was calculated long ago (A. I. Akhiezer?). This cross section is very small so one needs many-photon beams (laser beams) to collide in order to observe such rare events.

  7. Apr 26, 2009 #6
    How could you identify a single photon photon collision?
  8. Apr 26, 2009 #7
    Just as for any projectile-target scattering. The simplest way is to detect the photon deviation from its initial direction. No target, no deviation.

    Another effect is the polarization change. You know that a static magnetic field affects the photon polarization. The same effect will be produced in scattering of a high-frequency, short-wave polarized photons from a long-wave electromagnetic radio-wave with quasi-static magnetic field.

    Last edited: Apr 26, 2009
  9. Apr 26, 2009 #8
    Isn't that simply vector analysis? A Feynman diagram can be used for reflectivity in the same way and is a vector respresentation of the energy exchange. But can the exchange be isolated, that is, can two discrete photons be isolated and the resulting collison observed?
  10. Apr 26, 2009 #9
    Can a photon be considered to be the composition of an electron or a positron combined with a hadron or a lepton? Does a photon include the possibility of both a matter and antimatter waveform?
  11. Apr 26, 2009 #10
    I do not know, we have to compare the cross sections. If (one photon)-(one photon) cross section prevails and multi-photon one is much smaller, then yes.

    In case of Faraday effect mentioned above (one photon)-(multi photon) effect prevails.

  12. Apr 26, 2009 #11
    Do you mean why?

    Heisenberg's uncertainty principle?
  13. Apr 26, 2009 #12
    "What would a single photon scatter?" An electron, minus the energy needed to release it. If insufficient energy is present, once the uncertainty plus photon has sufficient energy an electron would release. By my understanding, sooner or later a single photon can cause scatter. Its not a mtter of if but when.
  14. Apr 27, 2009 #13
    as usual, out of my depth here, but i do not think you can have a photon-photon collision. in between the time a photon is emitted and the time it is detected/absorbed, a photon only exists as a probability - it is not a real entity at some specific place and time, it permeates the entire volume of the universe. perhaps i am misunderstanding what you guys are talking about...
  15. Apr 27, 2009 #14
    Put some dust in the volume and you will see the photon beam premises.

  16. Apr 27, 2009 #15
    thanks bob - i didnt go too far in the literature, but this is from wikipedia:
    "Two photons cannot ever collide.In fact light is quantized only when interaction with matter."

    i understand that photon beams (system) can affect each other gravitationally, but i thought the OP was aksing about individual photon collisions.
  17. Apr 27, 2009 #16


    User Avatar

    Staff: Mentor

    Indeed, two photons cannot interact directly (first-order Feynman diagram), but they can nevertheless interact indirectly via Delbrück scattering (a higher-order Feynman diagram).
  18. Apr 27, 2009 #17
    The gravitational photon beam interaction is negligible in comparison with the direct photon-photon interaction. The direct photon-photon scattering makes sense (exists) since the corresponding wave equations are non linear. It is a non-linear effect, if you like. The wikipedia statement is not correct.

    The Delbruk scattering is also a non linear effect - deviation of a photon in an external stationary strong electric field (near a nucleus) which itself is not a photon.

  19. Apr 27, 2009 #18
    Is Delbruck scattering where a photon collision results in an electron-positron annhilation which produces another photon?
  20. Apr 27, 2009 #19
    I looked it up and I was thinking of Pair Production but it describes a similar condition, photon collision in a nucleas field, energy > hv, electron-positron annhilation, and a photon is released (gamma I believe). If I remember correctly, one of the particles is moving in opposite time continuum to maintain symmetry. Must be different, the Feynman diagram doesn't reflect this.
  21. Apr 27, 2009 #20
    Um, doesn't the two-slit experiment embody direct photon-photon interactions?

    Of course the Feynman diagram utilizes other virtual particles, but two photons can cancel each other. No?

    Note: This is in response to the guy who said two photons cannot directly interact.
    Last edited: Apr 27, 2009
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?