[Wien] Questions about dielectric function of CNT.

[gbliu]

Dear Prof. Blaha and wien2k users,

       I calculated the dielectric function epsilon of (3,3) carbon 
nanotube (CNT).  I want to use the obtained epsilon to study the 
electromagnetic response of CNT. And what I concern is just the 
infrared~visible frequencies, i.e, the low energy part of epsilon, e.g. 
1~4eV. And I have some questions:

       (1).  Is my idea practicable? Can the epsilon obtained from 
first-principles method give enough precision (especially intraband 
contribution) which can be compared with experiments at low energy, say, 
1~4eV?

       (2).  The gamma for Drude terms is a user designated parameter. 
How to determine it reasonably?

       (3).  To check the first-principles epsilons, I compared results 
from wien2k and pwscf (a pseudopotential + plane wave package). Although 
the interband parts are similar, the intraband parts differ a lot. 
Wien2k gives out the plasma frequency of parallel polariztion 
omega_p=24.5eV, while pwscf gives out omega_p=3.57eV!  We know that 
intraband contribution is important at low energy. Which result is more 
reasonable? And is the calculated plasma frequency from first-principles 
methods reliable?

Best wishes,

Liu Gui-Bin
Dept. of Physics, The University of Hong Kong