So it is widely known that the width of the slit affects the interference pattern. And the double slit experiment is a demonstration of the wave nature of light. But i could casually see how the thickness of the slits would effect the particle nature of light, if we supposed them as the classical billiard balls.

But I looked around and could not find any information relating the effect of the thickness of the slits when they are particularly small. Thickness is defined in the attached photo.

My first thought would be to explore something that is the thickness on the order of the wavelength. But i know it would no longer be appropriate to think about it classically.

Image for clarityenter image description here


But I looked around and could not find any information relating the effect of the thickness of the slits when they are particularly small.

When the slits have small width (size in direction perpendicular to propagation), then the depth of the slit (size in direction of propagation) can have a pronounced effect on the intensity of output light. Namely, suppose the diaphragm with the slits is made of metal. Let incident light wave be polarized so that $\vec E$ vector is parallel to the slit, and let the wavelength be larger than twice the slit width. Then the wave wouldn't be able to be a standing wave between the slit walls, it would have to be "compressed" to have smaller wavelength to pass through still satisfying boundary conditions. This means that wave vector component in the direction perpendicular to propagation will be higher. The wave still has to maintain its frequency, so total wave vector magnitude must remain the same (because $|\vec k|=\omega/c$). Since (in two dimensions) we have

$$\vec k^2=k_x^2+k_y^2,$$

then to have $|k_x|>|\vec k|$, we must have $k_y^2<0$. This leads to an effect known as evanescent wave, where the wave, instead of oscillating like a sinusiod, falls off exponentially. So, as the depth of the thin slits increases, intensity of outgoing light will decrease exponentially.

If the wave is polarized in the direction parallel to slit width, there'll be no such effect, and the wave will pass virtually without change in amplitude, since boundary conditions on the metallic slit walls are "easier" in this case.

The shape of outgoing wave, unlike amplitude, won't change much: since the source is smaller than half wavelength, it's effectively a point source.

  • $\begingroup$ Ruslan, “polarized so that 𝐸⃗ vector is parallel to the slit”? You meant, that 𝐸⃗ is parallel to the slits width? $\endgroup$ – HolgerFiedler Apr 11 at 4:45
  • $\begingroup$ @HolgerFiedler no, I mean it's parallel to the symmetry direction (the direction in which the diaphragm is translationally invariant). $\endgroup$ – Ruslan Apr 11 at 4:51
  • $\begingroup$ So in the described by you first case 𝐸⃗ is parallel to the slits walls. $\endgroup$ – HolgerFiedler Apr 11 at 4:55
  • $\begingroup$ @HolgerFiedler yes $\endgroup$ – Ruslan Apr 11 at 6:00
  • $\begingroup$ Your implication is that the E-field is perpendicular to the width of the slit. So the relation of the size of slits width seems to be irrelevant. Maybe you would overread it again? $\endgroup$ – HolgerFiedler Apr 11 at 7:36

It will not affect the pattern but it may affect the quality. If it gets too thick you won’t see anything at all.


There are effects, but they depend on the light source. If the light source is very highly collimated (which implies that it also has high spatial coherence, light can pass through the slits with almost no scattering off the walls of the slits. Any light that does scatter off the walls will affect the interference fringe pattern downstream mostly by changing the fringe contrast.

  • $\begingroup$ @Ruslan, That is true! $\endgroup$ – S. McGrew Apr 10 at 19:28

Yes it does. But the pattern itself will not change in shape, or sharpness, only the intensity and visibility will change.

Let's start with thickness from very thick. What you will see is that the intensity of the pattern is decreased because of attenuation, and total internal reflection (less visible).

Now if the slits are not so thick, smaller then the wavelength, you can see the pattern well visible, with high intensity.

Now when the thickness reaches almost 0, you will have something called the plasmon.

  • $\begingroup$ See the image for clarity on my question. $\endgroup$ – Lambaste Apr 10 at 17:04
  • $\begingroup$ OK I edited. Still, there are effects. $\endgroup$ – Árpád Szendrei Apr 10 at 17:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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