In Drude-Lorentz equation which parameters are realistic for the consideration in spectroscopic ellipsometry. Any input is highly appreciable.
The pertinent parameters to be used depend on the material you want to analyze.
The formula you show includes a Drude contribution that is generally used to describe the response of free-carriers, and a sum or Lorentz oscillators that you can use to describe phonon absorption bands, interband transitions, etc.
If you want to analyze, for instance a silver material in the visible - IR, using the Drude contribution only might be sufficient.
If you want to analyze a non conducting polymer with absorption in the visible, you should skip the Drude contribution and use as many Lorentz terms as needed.
In the Lorentz term, woj stand for the spectral position, gammaj for the broadening, fj for the amplitude. Note that the Drude function is a Lorentz function peaking at a 0 frequency.
Dear sir, thank you very much for your valuable input and I am satisfied with the discussion you have presented. Our material of interest is Ti. Now we are keenly interested to know the realistic parameters in Drude-Lorentz equation for our case.
I did a test with a database dielectric function of polycristalline Ti, which is given from 0.6 eV to 6 eV; I managed to fit quite well with 7 oscillators, the main ones being 3 Lorentz at 1.56 eV, 3.1 eV and 0.7 eV, plus another one in the far IR (E->0) that I assume could be replaced by a Drude function. Likely, your material may have a different dielectric function (it might depend on fabrication conditions) so these values are just indicative. How do you fabricate the material?
The number of oscillators you shall use depends on your measurement, especially the spectral range. If you work in a reduced spectral range, probably you can reduce the number of oscillators.
Remind that all this analysis is totally phenomenological: if you fit perfectly your experimental spectrum in a selected range, you can be rather confident in the dielectric function you obtain and hopefully it will be useful for applications (for instance for multilayer engineering or so). However, as it is possible to fit almost any spectrum with different combinations and types of oscillators, getting "physical" info (such as the position of electronic transitions) from the fits is less straightforward.
- Badges
- Science topic
- Similar topics
- Mathematics