Hi Frederico,

About macroave, the first column refers to the z coordinates axis and the
second column refers to the value of the potential (planar average or
macroscopic average)?

I used the input(macroave-surface.in) of example folder.
If it that I wrote above is correct, then should I look to the second
column and see if the values are constant at region of the electrode and
the interface?

Comparing the bias cases of 0.0 and 0.3, I note that the values of the
case 0.0 is more constant than case 0.3(sometimes, varies of 0.4 between
two consecutive points ). Therefore, I must add more unit cell layers of
GNR in the scattering region?

About the contour option, I increased the value to 30,but the convergence
was not reached.

About your "PS", I didn't understand. My electrode is shaped by 3 unit
cell of GNR. Do you mean that when i want to increase de ribbon lenght i
can add only one unit cell? If is this, i know. I have included three
whole layers in ribbon(which is the electrode), because I thought it could
guarantee that the interface region would not suffer variation anymore,
but I was wrong. There are any recommendation? I am open to new ideas.

I want to thanks you Frederico, by your remarks.

Yoshi

> As you have already verified it is important to have a sufficient number
> of electrode layers in the scattering region, and the reason is for
> "things" to be sufficiently converged to bulk values in the electrodes'
> regions (as you probably know). Now, as you apply higher biases,
> depending on the screening properties of your material, you may need to
> include even more electrode layers. You can use the .VH file and the
> macroave program to see how the electrostatic potential is changing with
> the applied voltage. The applied bias should be screened inside the
> scattering region, resulting in an essentially constant (averaged)
> potential outside of it. If this is not the case, then you will have to
> include more GNR layers.
>
> As you seem to have a bias related problem, another thing to check are
> your contour options, mainly the number of points in the bias window.
>
> Best regards,
>
> Frederico.
>
> PS - You don't need to add 42 atoms each time you want to add more
> layers ... just to make sure.
>
>


Responder a