Page 1 of 1

### optical conductivity

Posted: Sat Mar 09, 2019 10:16 am
Hi all,

I want to whether there is a way to get conductivity /sigma from Yambo directly.
If answer is no, I can finish it by the relation between dielectric function and conductivity. Before that, I must know the unit of dielectric function in o. file. can anyone tell me this ?

Best,
Weiqing

### Re: optical conductivity

Posted: Sun Mar 10, 2019 11:54 am
Dear Weiqing,
no, there is not the possibility to have the conductivity as an output.
Note that the dielectric function by its definition is a mere number so it has not unit.
Best,
Daniele

### Re: optical conductivity

Posted: Mon Apr 01, 2019 9:38 am
Dear Daniele,
that the dielectric function by its definition is a mere number so it has not unit.
Actully, vaccum dielectric function has well-defined unit [(pi*e^2)/2h].
Can I understand in this way? In Yambo, the output of dielectric function is permittivity, the ratio of dielectric function to that of vaccum.

Best,
Weiqing

### Re: optical conductivity

Posted: Mon Apr 01, 2019 9:54 am
Dear Weiqing,
Can I understand in this way? In Yambo, the output of dielectric function is permittivity, the ratio of dielectric function to that of vaccum.
yes, it is as you said.

Daniele

### Re: optical conductivity

Posted: Thu May 30, 2019 4:48 am
Dear Daniele,

I want to make sure the dielectric function from Yambo is in the unit of vacuum dielectric function $\epsilon_{0}$, which can be rewritten as
$\epsilon_{0}=\frac {e^{2}}{2\alpha h c}$ where $e$ is the elementary charge, $h$ is exact Planck constant, $\alpha$ is the dimensionless fine-structure constant and $c$ is the light velocity. No any prefactor else?

I want to get conductivity, and a well-fined unit is very important. Now the problem I face is the conductivity I convert from Yambo is two orders of magnitude smaller than that of other articles, while other details are all same.

Best,
Weiqing

### Re: optical conductivity

Posted: Thu May 30, 2019 10:18 am
Dear Weiqing,
yes, the dielectric function is defined as
eps^-1=1+vX
where v is the Coulomb potential and X is the response function. In vacuum it is exactly 1.
If you are dealing with non-3D bulk systems (e.g. 2D systems etc.) issues related to the volume of the supercell arise.

Best,
Daniele