0
$\begingroup$

I am trying to understand the solution for electromagnetic plane waves in a conductor. I understand the derivation of the wave equation:

$$\nabla^2 \boldsymbol{\mathrm{E}} = \mu\epsilon \frac{\partial^2 \boldsymbol{\mathrm{E}}}{\partial t^2} + \mu\sigma\frac{\partial \boldsymbol{\mathrm{E}}}{\partial t} $$

Then what most authors immediately do is say it admits plane wave solutions with a complex wave vector $\boldsymbol{\tilde{k}}=\boldsymbol{k}+i\boldsymbol{\kappa}$. What I don't understand is why both the real and imaginary parts of this complex wave vector both have to be in the same direction. Why can't the wave be, for example, travelling in the $x$ direction, but decaying in the $z$ direction, like an evanescent wave?

$\endgroup$
4
  • $\begingroup$ They are not in the same direction unless $\boldsymbol{\kappa} = \lambda \boldsymbol{k}$ $\endgroup$
    – mostanes
    Commented Jan 9, 2020 at 16:37
  • $\begingroup$ Yes, the point I am making is that they do have to be in the same direction (i.e. $\boldsymbol{\kappa} = \lambda \boldsymbol{k}$), but I do not understand why. $\endgroup$ Commented Jan 9, 2020 at 16:42
  • $\begingroup$ In simple, nonmathematical terms, we expect that if a beam of light is propagating in a certain direction, and is also getting absorbed, then its intensity falls off in that direction -- not in some transverse direction. $\endgroup$
    – user4552
    Commented Jan 9, 2020 at 17:15
  • $\begingroup$ This does make sense, but then how come an evanescent wave for example decays in a direction different to its direction of propagation? $\endgroup$ Commented Jan 9, 2020 at 17:18

3 Answers 3

1
$\begingroup$

What other direction could the imaginary part point?

It’s not a facetious question.

With an (assumed) isotopic medium, there’s only one vector direction in the problem, which becomes $\hat{k}$.

To define another direction for the imaginary part, there has to be something physical to define that. An an-isotropy can do that. But absent that, the only available direction is along or against the only available vector: $\hat{k}$.

$\endgroup$
1
$\begingroup$

Try the general solution of the plane wave form $$ \textbf{E}=\textbf{E}_0 e^{i(\vec \lambda \cdot \vec r -\omega t)} \tag{1} $$ with \begin{align} \vec \lambda \cdot \vec r = \lambda_x x+\lambda_y y+\lambda_z z \end{align} and you will find that $\lambda$ must have a real and imaginary part. This follows because you have first and second order derivatives in $t$ so that one $t$-derivative of your ansatz (1) will return an imaginary multiple of $\textbf{E}$ while two derivatives in $t$ will return $\textbf{E}$ multiplied by a real number.

After all, the form of (1) is dictated by the assumption of a plane wave and by the general method of undetermined coefficients.

$\endgroup$
1
  • $\begingroup$ Thanks, I already did this, but it does not answer the question of why the imaginary part must be in the same direction as the real part. All this says is that there must be an imaginary part at all, which I agree with. $\endgroup$ Commented Jan 9, 2020 at 17:00
1
$\begingroup$

The reason is that you have assumed isotropic dielectric response. If you replace $\epsilon$ and $\sigma $ by tensors you can have the more general wave.

$\endgroup$
4
  • $\begingroup$ Is an evanescent wave not also an isotropic dielectric response? That occurs in an isotropic linear dielectric medium, and that wave decays in one direction while travelling in another. $\endgroup$ Commented Jan 9, 2020 at 17:18
  • $\begingroup$ Do you have total reflection in mind? $\endgroup$
    – my2cts
    Commented Jan 9, 2020 at 18:31
  • $\begingroup$ Yes, during total internal reflection when an evanescent wave travels along the boundary but decays into the medium. I know this is just for a normal dielectric rather than a conductor, but that is why I'm wondering what the difference is which means that in this case the decay and propagation are in the same direction. $\endgroup$ Commented Jan 9, 2020 at 19:35
  • 1
    $\begingroup$ Total internal reflection can be seen as due to anisotropy of the medium. $\endgroup$
    – my2cts
    Commented Jan 9, 2020 at 19:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.