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

Light wave/particle duality resolution

  1. Mar 27, 2010 #1
    Just two days ago I attended a lecture held by a professor at UiO who discussed the need to resolve the wave/particle duality model of light. He was unsatisfied by the deterministic approach to the problem (Bohr's principle).

    In general, he viewed the photon model as poorly defined and poorly founded, claiming the the historical experiments thought to have verified the photon theory could be explained with the wave model, with the exception of the light splitter experiment.

    If we would model the photon differently, not as a particle, but as a pulse of EM radiation lasting a time [tex]\tau[/tex] consisting of radiation with a period [tex]T[/tex] and an E-field [tex]E(t)[/tex], would it be possible to express the energy carried by the pulse using these three properties and fundamental constants in a way that satifsies previous measurements?

    I have thought about the idea, and I think the energy would have to be proportional to [tex]\tau[/tex] and the integral of [tex]E^2(t)[/tex] over the pulse and inversely proportional to [tex]T[/tex] (known from previous measurements). I also consider [tex]c[/tex], [tex]\epsilon_0[/tex] and [tex]h[/tex] as neccesary constants, but I have been unable to reach units of joules using combinations of these.

    Does anyone have any thought on the matter?
  2. jcsd
  3. Mar 28, 2010 #2
    I went to the University of Oslo website to check out the physics department and see if I could guess who might have put forward this idea, but couldn't. I got interested in this subject years ago when I figured out how much sense it made to explain the photoelectric effect with the wave theory, after hearing so many times that it was impossible.
  4. Mar 28, 2010 #3
    The person in question was, I think, the first amanuesis, who researches quantum optics and biophysics. I think his name is Vistnes, but I can't remember precisely.
  5. Mar 29, 2010 #4

    We know that EM waves have an aspect of radiation pressure, but isn't the essential problem of modeling a photon entirely as a wave that the wave can't convey the amount of momentum that the particle is shown to have? (Photons have a localized momentum when colliding with other particles in the Compton effect)
  6. Mar 29, 2010 #5
    The wave explanation of the Compton effect is particularly satisfying. Light interacts with electrons in the COM frame when they both have the same wavelength. (And in QM wavelength equals momentum). The superposition of the incident and recoiling electons sets up parallel sheets of charge exactly one-half wavelength apart (just like the well-known solutions of the 1-d potential well) and as is well known in classical electromagnetics, there is a very strong reflection from a structure of that kind.
  7. Mar 29, 2010 #6


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Would that 'wave theory' still hold water in trying to describe (i) angle-resolved photoemission, (ii) resonant photoemission, and (iii) multiphoton photoemission?

    I've also heard so many times of people claiming that such wave theory can explain the naive photoelectric effect, but never have these people attempted to extend such theory to cover the more detailed aspect of photoemission phenomenon. Such details are what made a cow not a sphere when one looks at it with finer instruments.

  8. Mar 30, 2010 #7
    Something about that statement triggered a sense that it could very easily lead to incorrect assumptions. And after a little reflection I realized how deceptive the equation is:
    [tex]\rho = h \omega[/tex]

    There the particle's momentum is compared with the wave's wave number. In other words, properties pertaining to 2 separate objects are being compared. In de Broglie's text book "An Introduction to the Study of Wave Mechanics" he specifically states that the wave length or wave number is that of the "associated wave".

    On page 6, de Broglie gives a pretty convincing argument about why a particle can't consist of a wave packet, at least the same wave packet as the "associated wave". When a particle-wave pair in transit encounters a highly dispersive obstacle, the wave will be dispersed and eventually disappear from observation. Yet the particle will not disappear and cannot be destroyed in that situation.
  9. Mar 30, 2010 #8
    Quantum mechanics always seems to present us with these little conundrums in pairs. Yes, the integrity of the wave function is problematic, but so is the question as to how the particle goes through two slits at the same time. We can't let these little puzzles stop us from using the best methods as they apply to specific problems. In any case, in the traditional Compton scattering experiments, we are just looking at the angular distribution of different scattered light frequencies. We aren't ever really tracking individual electrons. And the wave calculation gives us the same basic result as the familiar billiard-ball calculation.
  10. Mar 30, 2010 #9


    User Avatar
    Science Advisor

    Well, how would this model explain the photon antibunching seen in experiments where emission from a single photon source is directed towards two detectors using a beam splitter and the detections are always anticorrelated? Contrary to common belief, this is THE smoking gun experiment for the particle nature of light.
  11. Mar 30, 2010 #10
    I am aware of this. It is as far as I know (according to the professor who held th lecture) the single reason for keeping the photon concept. (Of course, he's biased.)
  12. Mar 30, 2010 #11


    User Avatar
    Science Advisor

    I see your point. However, I agree with ZapperZ. While it is true that the experiments which were historically considered to be evidence for the particle nature of light - like the photoelectric effect - can be described in a wave picture, that does not mean that these models can be applied to all complicated forms of photoemission experiments.

    Besides that: Even if photon antibunching was the only reason to keep the photon picture, it would still be a good one.
  13. Mar 30, 2010 #12
    Of course. I still beleive that a future model of light will be primarily wavelike, even though I do not know how issues like antibouncing and various photoemission phenomenon. It might be likely that a new notion of "electromagnetic wave" is neccesary.
  14. Mar 30, 2010 #13


    User Avatar
    Science Advisor

    The current quantum optical model of light is already pretty wavelike. The photon aspect just means that interactions take place at discrete positions and in discrete chunks of energy. The remaining part of the physics involved is described in terms of probability amplitudes (just like almost everything else in physics) which are indeed "wavey". The times of photons being interpreted as tiny balls are (fortunately) long gone.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook