TranSIESTA: low current values with 4.0

Asked by Tristan Zaborniak on 2019-10-10

I've been attempting to construct an I-V curve using the 4.0 version of the Siesta bundle for the structure whose .fdfs I've attached via the dropbox link below, by first running the electrode calculations, and then running the scattering region Transiesta/TBTrans calculations (at select biases). This structure I've labelled as 'extended'.

All calculations run through to completion, but the current values output are extremely low (on the order of 10^-30 A). In running a similar structure, the only difference being that the electrodes were entirely bulk (or crystalline, if you will), such that the sulfurs at either end were bonded to bulk gold instead of linear gold chains (which then attach to bulk gold), the current values output were much more reasonable (on the order of 10^-7 A). This structure I've labelled as 'compact'.

My question is: why? Between these structures I did not change any input parameters but the lattice vector in the transport direction. Previously I've ran structures similar in setup (with the bulk electrodes transitioning to a straight chain of gold atoms) and not faced the tiny current issue.

Here is the link to the structures: https://www.dropbox.com/sh/62f45wg0bjm43d1/AABX5jKLVqb97A3Dr4VAVlIXa?dl=0

I'd appreciate your comments on this.

Question information

Language:
English Edit question
Status:
Answered
For:
Siesta Edit question
Assignee:
No assignee Edit question
Last query:
2019-10-10
Last reply:
2019-10-11
Nick Papior (nickpapior) said : #1

You haven't converged k-points in either SCF or transport calculations (at least from your fdf files). In the tutorial I linked in your previous question there is detailed explanations of why one needs to converge k-points.

Also, your two structures are *very* different. You seem to indicate that the linear chain shouldn't make a difference? But it will.

Can you help with this problem?

Provide an answer of your own, or ask Tristan Zaborniak for more information if necessary.

To post a message you must log in.