1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Photon polarization question

  1. Dec 19, 2008 #1
    1. The problem statement, all variables and given/known data

    Hi, I have a question regarding a set of equations in Feynmann volume 3. On chapter 11, page 11, Feynmann discusses the right-hand and left-hand circulation for the polarization of the photon. He states: "In the classical theory, right-hand circular polarization has equal components in x and y which are 90 out of phase. In the quantum theory, a right-hand circularly polarized (RHC) photon has equal amplitudes to be polarized |x> or |y>, and the amplitudes are 90 out of phase. Calling a RHC photon a state |R> and a LHC photon a state |L>, we can write:

    2. Relevant equations

    |R> = (|x> + i|y>) / sqrt(2)
    |L> = -(|x> - i|y>) / sqrt(2)."

    3. The attempt at a solution

    This is more of a question, but isn't it true that for RHC the x-component is 90 ahead rather than behind? Thus, shouldn't we place the i (which is a 90 phase factor) in front of the |x> instead of the |y>? Same question goes for the LHC equation, but with phases reversed. Thanks.
     
  2. jcsd
  3. Dec 21, 2008 #2
    Sorry if the question was not made clear; let's rephrase it. In RHC photon polarization, it is the x-component of the electric field vector that is 90 degrees, or exp(i * pi / 2) = i phase ahead. The complete vector (from the sum of the components) rotates in a counterclockwise fashion, hence the right hand orientation with the thumb pointing out of the screen.
    However, when we look at the equation for the state |R> in Feynmann, we get, as before,

    |R> = (|x> + i|y>) / sqrt(2).

    Here is the problem. Take the amplitude to be x-polarized and the amplitude to be y-polarized:

    <x|R> = 1 / sqrt(2)
    <y|R> = i / sqrt(2)
    Therefore, <y|R> = i<x|R>.

    This equation shows the y-component to be 90 degrees ahead rather than the x-component. So, the question is, is the equation wrong and should we instead put the i in front of the |x> in the equation for |R>? Or is it really the case that I am overlooking something? Given that this is the Feynmann lectures on physics, the latter is probably the case...

    Finally, I have to ask as I'm new here, is this the right place to post a question such as this one, or should I ask in a different area? Thanks for all your time.
     
  4. Dec 23, 2008 #3

    Redbelly98

    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    Welcome to PF :smile:

    I've been thinking about your question, and finally this morning wrote out the equations for Feynman's |R> state,

    |x> + i|y>​

    in terms of sines and cosines. I get a left-handed polarized wave! I wonder if Feynman was defining RH and LH polarizations differently than us.

    I have to leave for work soon, and after today I'll be traveling for the holidays. I'll try to post back here on Saturday, with more details. Feel free to post a VM* as a reminder if I haven't done that by Saturday evening (Eastern USA time).

    VM = Visitor Message, on my profile page:
    https://www.physicsforums.com/member.php?u=122961

    Regards,

    Mark

    EDIT:
    p.s., sorry about the delayed response. This one is a little more involved than most questions in here.
     
  5. Dec 24, 2008 #4

    Redbelly98

    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    Okay, here is what I have done.

    I write the electric fields of the |x> and |y> states as follows. I'm leaving out the normalization factors and E-field amplitude, since those don't affect the orientation or shape of the fields.

    |x> → Ex = Re { exp[i(kz-ωt)] } = cos(kz-ωt)
    |y> → Ey = Re { exp[i(kz-ωt)] } = cos(kz-ωt)​
    and
    i|y> → Ey = Re { i exp[i(kz-ωt)] } = -sin(kz-ωt)

    Combining
    |R> = |x> + i|y>​
    gives
    Ex = cos(kz-ωt)
    Ey = -sin(kz-ωt)​

    This describes a left-hand circularly polarized wave, i.e. the E-field resembles a left-handed screw.

    However, if you point your right thumb in the direction of propagation (+z), your right-hand fingers will curl in the direction that the E-field rotates at a stationary point. I am wondering if Feynman had this in mind for right-handed circular polarization?
     
  6. Dec 25, 2008 #5
    Here is how RHC is defined in volume 1, chapter 33, page 2: "If the end of the electric vector, when we look at it as the light comes straight toward us, goes around in a counterclockwise direction, we call it right-hand circular polarization." So apparently we are not viewing the light from the origin as it travels away from us in the positive-z direction; rather, we are viewing the light coming towards us.

    Even Feynman himself admits that in RHC, the y-component of E is -i ahead, or, in other words, i behind the x-component. So we are stuck with the definition we first were thinking of. Let's then go back to the equation in my second post:

    <y|R> = i<x|R>.

    Here's a wild guess: We would think that <y|R> corresponds to Ey. Well, perhaps not. Maybe <R|y> is really what works. This is a possibility, because if we take the complex conjugate of each side, we get:

    <R|y> = -i<R|x>,

    which is what we are hoping for. I doubt <R|y> really gives the y-component, but it's a possibility...

    Well redbelly, let me know what you think about this way of looking at it. If, in fact, <R|y> is the correct representation, then it would be great to know why this is the case.
     
  7. Dec 28, 2008 #6
    Okay guys, I talked with a friend of a friend (he's actually a PhD in physics) and have come up with the answer, so here we are. If we look at an RHC photon (using Feynmann's convention) coming towards us, then the x-component will be 90 degrees ahead, given our x-y plane. But, if we look at it coming away from us, then, given the mirror image of this x-y plane, the y-component will be 90 degrees ahead. So we may use either coordinate system and put the i either in front of the |x> or the |y>.

    Ultimately, however, it does not matter. Physics is not described by amplitudes; rather, it is described by probabilities, which are the absolute squares of the amplitudes. In fact:

    P(x-polarization) = |<x|R>|^2 = (1 / sqrt(2))^2 = 1 / 2,
    P(y-polarization) = |<y|R>|^2 = (i / sqrt(2)) * (-i / sqrt(2)) = 1 / 2.

    So really, we have a 50-50 chance of polarization in both directions and it ultimately does not matter which coordinate system we use, whether it be viewing an x-y plane as the light comes towards us or an x-y plane as the light moves away from us.
     
  8. Dec 29, 2008 #7

    Redbelly98

    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    Ah, there it is. RHC is defined oppositely in the well-known textbook Optics by Hecht and Zajak (or by just Hecht, depending on what edition it is). I just checked there, and they say E rotates clockwise for RHC coming towards the observer.

    I've always used the Hecht and Zajak definition, and must say I prefer it because the spiraling E-field vector in an RHC wave has the same sense as the screw threads in a right-handed screw.

    But if somebody like Feynman wants to define it differently, who am I to argue? :smile:

    Regards,

    Mark
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Photon polarization question
  1. Polarization question (Replies: 2)

  2. Photon question (Replies: 7)

  3. Polarization Question (Replies: 10)

  4. Polarization question (Replies: 1)

Loading...