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

Bending light with a magnet

  1. Jan 3, 2009 #1
    I was wondering if it was possible to bend a beam of light with a magnet. If this is possible could someone tell me what the requirements for the magnet be to do such a thing.
  2. jcsd
  3. Jan 3, 2009 #2
    The following is from: http://www.wonderquest.com/extinctions-safetyglass-magnetslasers.htm

    Q: Can a magnetic field interrupt the path of a laser? Can a magnet bend light? —"Bending light", Vancouver, British Columbia, Canada

    [NASA] Laser light reflected by mirrors but unbent by magnets

    A: A magnet can bend the path of any moving charged particle. In fact, that’s how your TV screen displays a video picture. A magnet deflects a beam of electrons to create a video pattern on the screen. Light, however, has no charge and therefore its path is unaffected by a magnet.

    But, you mention a magnetic field, which includes changing fields. If it’s a changing field, things get more complicated. Any changing magnetic field generates a changing electric field and that produces an electromagnetic wave.

    Electromagnetic (EM) waves cannot interact directly with light photons since photons have no charge. EM waves do not bend light, at least enough that we can measure. If radio waves, for example, bent light appreciably then a transmitting radio station would look blurry. But stations don’t go blurry.

    Actually, electromagnetic waves can bend light through an indirect, quantum effect—but to such a tiny degree that we cannot measure it. This quantum effect (called Delbrück scattering) "is a process where, for a short time, the photon disintegrates into an electron and positron pair," says Norbert Dragon, physicist at the Institute for Theoretical Physics in Hanover, Germany. The charged pair interacts with an EM wave and then recombines into the photon with a changed direction. Thus, the EM wave bends the light.

    "More probably the charged pair will annihilate into two or more photons—this process has been observed under extreme conditions—but, then, the light ray is not bent but rather split into several rays," says Dragon.
  4. Jan 3, 2009 #3
    Darn. There goes my invisibility cloak lol. The only other option to bend light around myself would be gravity, but Im sure a suit like that would have some major consequencies(everything flying at it lol). Oh well, thx for the reply :)
  5. Jan 3, 2009 #4
    I'm probably mixing theories, but can't photons interact with EM waves, seeing as EM waves *are* photons, and photons, being bosons, tend to clump into similar states?

    If you have a standard bar magnet and an old CRT television or computer screen, try it for yourself. As you bring the magnet near the screen, you'll see the picture go crazy. This is due to the force of the magnet on the stream of electrons through the television. In fact, bending light with magnets is the fundamental principle behind CRT televisions!
  6. Jan 3, 2009 #5
    Does that mean that because light has no mass it is not affected by gravity?
  7. Jan 9, 2009 #6
  8. Jan 9, 2009 #7


    User Avatar
    Science Advisor

    No, that quote has nothing to do with either mass or gravity. And, in fact, while it is true that light has no mass, it is affected by gravity. The fact that light bends around stars was one of the first important tests of relativity.
  9. Jan 9, 2009 #8
    Eddington 1919 I presume? Bending of starlight by the the sun 1,75 arcseconds nearest to the Sun. By comparing starcharts of 1 the stars during the eclipse and 2 at a moment that sun and moon are enough away (mostly I say half a year later) it seems that the stars have moved away from the sun. Explicitly not toward the Sun! In our conceptual scheme we think light going straight...
  10. Jan 12, 2009 #9
    This is a bit of a weird statement (sorry)... First you say correctly that CRT tv's bend a stream of electrons with magnets. Then you conclude that they are bending light with magnets?

    CRT tv's are not bending the path of light with magnets, they are bending the path of electrons.

    I assume you meant electrons in your last sentence? (But if you do, the post would not really be relevant in the topic of bending light with magnets...?)
  11. Jan 12, 2009 #10
    This is very true. I must have misread X-(
  12. Jan 12, 2009 #11
    Photons have no restmass yet bend because of their relativistic mass = hf/c^2. Have sought for some construction like: no rest charge but yet relativistic charge, but I was assured that no charge at all is associated with photons...

    Yet perhaps GRAVITOMAGNETISM yields anything for your problem?
  13. Mar 28, 2009 #12
    The last time I quoted this something went wrong. I mentioned "the mexican hound" The effect that stations actually used to get blurry
    when received with old radio's. This effect has been corrected by the automatic frequency control: a way to let the receiving radio follow the base frequency of the carrier wave.
    I still don't have an answer to this remark Pallidin, stations used to get blurry before this invention!
    greeting Janm
  14. Mar 29, 2009 #13
    I understand what your saying, but read the statement again.
    It's saying that the physical radio tower does not get visually blurry while emitting large amounts of EM radiation. Thus it does not distort light.
    It's NOT talking about a "blurry" received radio signal.
  15. Mar 30, 2009 #14


    User Avatar

    One effect a magnet has on a photon is rotate it's plane of polarization (must be very strong). Maybe u could use that for your invisibility cloak.
  16. Mar 30, 2009 #15


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    We have not yet seen any photon-photon interaction, which is one of the higher-order interactions predicted within QED. For such a thing to have an appreciable probability to occur, we need very high energy photons (gamma-gamma interaction). Without that, the probability of such interaction is minuscule to non-existent.

    Coming back to this thread, we need to make sure we FRAME the scope/range of this question properly. It is asking for the typical classical E&M scenario within the typical working range that we can achieve under ordinary condition. So let's not start messing about with exotic circumstances, tempting as they may be. Magnetic fields do not bend light under a typical and large range of parameters. Period.

  17. Mar 30, 2009 #16
    Hallo ZapperZ,
    So if people as important as Eddington and/or Einstein state that light bends by gravitation that is to be taken seriously and this question that it could possibly be the magnetic field of the Sun which could be the cause for this bending has to be closed?
    I might say magnetic field is a better candidate for bending light then gravitation!
    greetings Janm
  18. Mar 30, 2009 #17


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    What does "light being bent by gravitational field" has anything to do with "light being bent by magnetic field"?

    It really is a simple question. It doesn't involved higher-order QED interactions, nor does it involve exotic particles such as axions. If it does, it will be in either BTSM forum, or even HEP forum.

    The classical Maxwell Equation provides no mechanism for light to be bent in a magnetic field. If you know of a peer-reviewed publication that has something to the contrary, please cite it, and then, we'll open that discussion.

  19. Mar 30, 2009 #18

    See the English translation of the original paper by Euler and Heisenberg:


    In the low energy limit, there are small nonlinear corrections to Maxwell equations. If you have a magnetic field in a region and an electromagnetic wave moves through thast reason, then theeffect of the magnetic field is the same as if the vacuum were a so-called birefringent medium. I.e. the medium has an index of refraction that depends on the polarization of the electromagnetic wave relative to the magnetic field orientation.

    Details are given here:


    Now, the fact that you have an effective index of refraction, means that the light will indeed be deflected by a magnetic field. This doesn't mean that you can measure it. If you take a laser and shine it through the field of an ordinary magnet with a field strength of 1 tesla, then you should take into account that the photons emitted by the laser do not have an infinitely sharply defined momentum.

    So, if they are deflected in some direction, a miniscule amount of mometum is added to a state of which the momentum was spread out quite widely. The final state and the initial state are then almost identical as far as the momentum is concerned. According to quantum mechanics, you then cannot detect that any momentum has been added at all, at least not in a case by case basis. You would have to average over a huge number of photons to see a tiny effect. And so far we haven't even considered the effect of noise...

    In case of so-called mnagnetars which are neutron stars that have field strengts of billions of teslas, the effect can be more easily detected. But so far it hasn't.
  20. Mar 30, 2009 #19
    ZapperZ, but what we call "classical electromagnetism" is in reality quite complicated quantum mechanics involving coherent states of photons.

    Why can't we call the Euler-Heisenberg lagrangian classical too? I mean, if the nonlinear corrections had been larger (e.g. if the fne structure constant had been larger or electron mass smaller), then Faraday could have measured a vacuum birefringence effect and Maxwell's would have been able to put everything together and write down the Euler-Heisenberg equations. The coefficients of the nonlinear terms would then be phenomenological parameters.

    Euler and Heisenberg would then later explain the origin of these terms. But everyone would refer to the Euler Heisenberg equation as the "classical Maxwell eqations".
  21. Mar 30, 2009 #20
    Over wide ranges of conditions photons (electromagnetic waves) do not interact....but gravitational waves do interact...to say it another way, photons don't interact, gravitons do....hence the formulation of gravity in general relativity is considerably more difficult than in Maxwell's electromagnetic fields....Einsteins tensor is more complex than its counterpart in EM.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook