Restrahlen Effect and Plasma Frequency

[Master J]

If we model the electrons in a metal as free, we can get nice expressions for the relative permittivity and hence reflectance. Above the plasma frequency (usually UV), the reflectance falls from about 100% and metals become transparent.

We can also develop expressions for relative permittivity from a simple phonon model. We seen then that between the resonant Transverse Optic and Longitudinal Optic frequencies, the reflectance of the material is again about 100%, and drops dramatically after that, being low, before the TO freq. also.


My question is, can both models be applied to a metal? Clearly only the second works for dielectrics. But metals have phonons too.

Does the plasma frequency, where the reflectivity drops dramatically in a metal, correspond to the LO frequency at the end of the Restrahlen band in the phonon model?

My gut feeling is that these are 2 separate models, and the plasma one is by far the more dominant in metals.

Can anyone enlighten me on this?

 

[Master J]

I think I may actually have found the answer.

Light only couples to charged atoms...hence the phonon model works with ionic materials, or materials that have some polar character.

But in metals, we have a completely different type of bonding, where electrons are delocalized throughout the metal, and the bonding has no polar character. Hence metals have no optically active phonons (pure metals, tho I guess certain compounds etc MAY have).

Is this correct?