Dear Eliya Asmani,
This is because, in SIESTA, the dielectric function is reciprocal to the
volume of the supercell. So, you need to use the same supercell. Or you
can multiply the volume of the supercell for each job and then divided
by a same volume factor at last.
Kind regards,
/Guangping Zhang
On 2015/5/4 13:16, Eliya Asmani wrote:
Dear all
I calculated imaginary part of dielectric function for a metal cluster
with different radius 0.5, 1 and 1.5 nm. As you know in the siesta for
open system, supercell should be determined. My results are dependent
on size of supercell and with increasing radius of cluster, size of
supercell increases and value of dielectric function decreases while I
don't expect this behaviour.
What is the unit of permittivity in siesta I think it isn't dimensionless.
I appreciate if anyone has the same experience help me to solve this
problem.
Regards