Chemical Potential Input Options

Asked by Adam Bruce on 2017-08-08

The manual gives an example of the TS.ChemPot block as

%block TS.ChemPot.Left
mu V/2
contour.eq
begin
C-Left
T-Left
end
%endblock

using bias range of -V/2 to V/2. Would changing the range to 0 to V have an effect on the Contour blocks? For instance, if I set

%block TS.ChemPot.Left
mu 0.1
contour.eq
begin
C-Left
T-Left
end
%endblock

%block TS.ChemPot.Right
mu 0.0
contour.eq
begin
C-Right
T-Right
end
%endblock

%block TBT.Contour.line
from -0.1 eV to 0.1 eV
delta 0.001 eV
method mid-rule
%endblock TBT.Contour.line

would this yield a difference integral than if I used a bias range of -.05 to .05?

Question information

Language:
English Edit question
Status:
Answered
For:
Siesta Edit question
Assignee:
No assignee Edit question
Last query:
2017-08-08
Last reply:
2017-08-08
Nick Papior (nickpapior) said : #1

No, there will be no difference (when using TS -> TBtrans).

This is because the bias applied is self-consistently solved and each of the electrodes are considered "in equilibrium". I.e. when you apply _any_ V to any electrode you merely lift the electronic structure of the electrode accordingly.
This is a good exercise to assert.

So, essentially you are free to set the applied bias' as you please. But, for two terminal cases you are encouraged to maintain the -V/2 -- V/2 scenario due to the following:

When TBtrans calculates transport, DOS, etc. the energy reference is the "Fermi level" of the non-bias system. I.e. if you apply 0 -- V you will experience a change of reference energy by V/2 because the self-consistency will lift/reduce all states so that charge neutrality is retained.

If this isn't clear, please perform _any_ 2-terminal bias calculation with -V/2 -- V/2 and 0 -- V and compare the two graphs. The latter should be shifted by V/2. And for practical cases it becomes much more convenient to retain the -V/2 -- V/2 scheme because then you do not have to "correct" the energies to compare different applied bias'.

Also, your mentioned blocks are not consistent with a 0 -- V bias, they should look like this:

%block TS.ChemPot.Left
mu V
...
%endblock

%block TS.ChemPot.Right
mu 0. eV
...
%endblock

And importantly, all energy contours should also be shifted accordingly (in case you want 0.05 on both sides of the chemical potentials):
%block TBT.Contour.line
from -0.05 eV to V + 0.05 eV
delta 0.001 eV
method mid-rule
%endblock TBT.Contour.line

Note that transiesta is not going to handle if you apply a negative bias with the above input. This is due to lack of implementation, but again, the end results are identical so no need to introduce such difficulty.

Can you help with this problem?

Provide an answer of your own, or ask Adam Bruce for more information if necessary.

To post a message you must log in.