[DFTB-Plus-User] R: DFTB+NEGF with a gate contact

Mon Mar 18 19:19:19 CET 2013

Hello,

The gate direction should be general now. I also checked the code. In case
download the latest release.

Direction is specified by GateDirection = 1 2 or 3  (1==x, 2==y, 3==z)

GateDistance controls the distance of the gate from the central axis of the
Poisson Box. 

Therefore it should be consistent with the numbers used in
PoissonBox=(Lx,Ly,Lz)

In practice GateDistance < Ly/2   (for a gate in y direction)

Remember that default units are [AtomicUnits] and not [Angstrom].

Alex.

Da: dftb-plus-user-bounces at dftb-plus.info
[mailto:dftb-plus-user-bounces at dftb-plus.info] Per conto di Yuranan
Hanlumyuang
Inviato: mercoledì 13 marzo 2013 15.25
A: dftb-plus-user at dftb-plus.info
Oggetto: [DFTB-Plus-User] DFTB+NEGF with a gate contact

Hi,
   I have some basic questions about NEGF calculations with three contacts,
using the Poisson box

(1) Why is in the code, the bias direction is set to the x-direction. In the
manual, it's suggested that
the transport direction is the z-direction, while the gate is y-dir.
However, in the  source code  poisson.F90 
where the gate  direction is  automatically set to y-dir, and the bias is
set to x-dir.

(2) What are the criteria for the gate distance (i.e. Rmin_Gate)? Sometimes,
I got a warning message
that "WARNING: Gate Distance too large!". I have looked around the source
code, but still couldn't figured
out which section  decides the value of this parameter.

(3) I've missed the the discussion on the bugs of DFTB+NEGF version r4198,
and how they
are fixed in version r4305. Is version r4198 containing bugs on the
three-contacts sections of the code?
If you could repost the old threads, I would be really appreciated. I am
using version r4198...

Yuranan Hanlumyuang
University of Houston

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20130318/c7ee6da4/attachment.htm>