large oscillator strength but small imaginary dielectric function

[awang95]

Hi,

I have computed the following spectra, but I'm wondering, specifically for the peak around 1000 nm, how it's possible to have such a large oscillator strength when the absorption coefficient (as given by the imaginary dielectric function) is small as shown. I have some ideas in my mind but am really not sure (does the oscillator strength include IR transitions while the imaginary dielectric function only includes electronic transitions?) - any help or thoughts on this would be really appreciated; also attached are the relevant job files

thanks!

pic.PNG

[svijay]

Dear awang95,

Please provide relevant output files (such as the OUTCAR) in accordance with the forum posting guidelines here: https://www.vasp.at/forum/viewtopic.php?t=17928. It would also be beneficial if you could provide a simple script that you use to plot the quantities in your plot from the relevant output files; thank you!

Sudarshan

[awang95]

sorry about that - here are the missing files that are also relevant

[kaynat_kalvi]

Hi, I am stuck that the same understanding issue, if you have resolved your's please explain it to me as well.

[alexey.tal]

Dear kaynat_kalvi,

In vasprun.xml the components of the dielectric tensor are written separately, but the oscillator strength values are summed up together for xx, yy, and zz.
Thus, some of these transitions are only present for one component but are not visible in the others.
Here is a plot for the dielectric function and transitions provided by awang95. For example, the transitions at around 1000 nm are only present in the epsilon_xx.

Best wishes,
Alexey