How Does Thin-Film Interference Create Color Effects?

[austinv]

I consider myself very well-read on the phenomenon of thin-film interference: how it works, how it's used in lens making, etc. HOWEVER... there's one thing I don't get:

Thin-film interference only occurs for one specific wavelength at any given point (according to all the sources I've read), not one specific color (a range of wavelengths), just one individual wavelength. This makes total sense based on how the phenomenon occurs, except that when I think about it, this should mean that thin films would not reflect vibrant colors.

Let's say we're dealing with a thin film that has an even thickness, and only one specific wavelength is affected destructively, and another constructively. If white light shines on it, we should not see a vibrant color, we should see the white light with less of the wavelength that destructively interferes and more of the one that constructively interferes... but white light with less of one wavelength and more of another will still look like white light because we're still seeing all the other wavelengths reflected back to us. Thin film reflections should not seems colored for the same reason that white light emitted from a star seems white even though spectroscopy reveals many missing wavelengths in its spectrum.

Please help me (and hopefully others as well) understand how it's possible for thin films to create colors.

And on a related note, how do anti-reflection coatings work, even multi-coated ones, considering each layer is only effective for one wavelength and visible light (400-700nm) is at least 300 different wavelengths?

Thank you so, so much!

 

[Vorde]

Feynman's QED treats this problem really well. One thing I would say is that you are wrong about the white-color aspect. When light of all sorts of different frequencies enters the eye, all of the color-receiving nodes are activated and our brains perceive white light. When your eyes receive light that has more of one wavelength than another, you will perceive a color that is whitish, but resembles the color of the primary frequency (I believe this is complicated by how our brains perceive colors, but the gist is correct).

With that in mind, consider a thin film (soap bubbles are the example used by Feynman) that you are shining white light on. Every frequency of light will undergo some sort of interference, but because the wavelengths of the various photons are different, the resulting interference pattern will be different from one frequency photon to another.

The result of this is you have some patches with a lot of one color (where that color constructively interfered) and a little of another color (where that color destructively interfered). When this happens you won't see white anymore but (primarily) the color that constructively interfered.

In a circumstance where there are lots of different colors interfering (white light), you get all sorts of spots where various colors mix in differing amounts to produce the rainbow-esque color effects you see.

 

[Emilyjoint]

when you look at different angles the pathe through the film is different and therefore a different colour will interfere to give a max