0
$\begingroup$

The reflection of the interface of two layers is described by Fresnel's equations. For instance, the reflectance of an Al bulk at 1 micron is around 0.9, this is the reflectance of the interface between air and Al, and it shouldn't change no matter how thick the Al layer is, as long as the interface is still the same. Now assuming we polish the Al to a 1 nm thin film from the other side, and keep the air-Al interface intact. Then the reflection can be lower than 0.9. The refractive index of Al doesn't change, and the smoothness of the interface is still the same, why the reflectance can be different?

$\endgroup$
  • $\begingroup$ The construction of the Fresnel equations relies on the assumption that the polarization surface current at the interface has zero thickness, but this is obviously just an approximation. If your medium is thinner than this surface current (i.e., as akhmeteli points out, thinner than the skin depth) then the Fresnel equations are at least susceptible to change. $\endgroup$ – Emilio Pisanty Jun 13 at 1:44
  • $\begingroup$ @EmilioPisanty : I think another assumption one uses to derive Fresnel formulas is important: that there is no wave* falling on the surface in the direction opposite to that of the incident wave. Such wave* is important for a layer of finite thickness. $\endgroup$ – akhmeteli Jun 13 at 9:50
0
$\begingroup$

I believe 1 nm is of the order of magnitude of the skin depth for visible light in aluminum, so part of light gets to the farther surface of the Al layer and partly exits the layer, so the reflectance decreases. Remember that reflectance typically depends on the thickness of the layer.

$\endgroup$
  • $\begingroup$ Thank you very much for the information. Are there any references you know talking about how to calculate the reflectance and transmittance in this case? I've tried to find them online but I got nothing... $\endgroup$ – user221757 Jun 19 at 19:01
  • $\begingroup$ @user221757 : apps.dtic.mil/dtic/tr/fulltext/u2/a098940.pdf seems relevant. It's quite doable, but not a walk in the park. $\endgroup$ – akhmeteli Jun 20 at 1:48

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.