[DFTB-Plus-User] Transport calculation problem for defining the geometry

[Gabriele Penazzi]

On 01/17/2013 02:11 PM, bedamani singh wrote:
> dear DFTB users,
> I have optimized a whole system (device and two contacts having two
> principal layers) and then tried for the transport calculation. But an
> error "contact 1 is not parallel with any of the axis" comes up. 
> Since the optimized structure may or may not be perfectly symmetric,
> the two principal layers in the contacts may or may not be the exact
> replica of each other. Hence for the transport calculations, is it
> necessary to optimize the system as a whole? Or can we optimize the
> contacts separately and afterwards join with the device?   
>
>
> I have tried with the following input files (dftb_in.hsd and geo.gen),
> and it always gives error like
> "-> Contact 1 does not consist of two rigidly shifted layers"

Dear bedamani,

the contact principal layers should be an exact replica of each other,
otherwise the assumption behind the calculation of surface green's
function do not hold. You can override the requested value for the
tolerance using the flag ShiftAccuracy (see manual) to force the code to
ignore a non negligible mismatch. You need to specify the flag in every
contact block. However, the numerical error will result in a slightly
wrong transmission (probably you won't get ideally sharp transmission
steps).

I see that your structure is an ideal GNR. In this case I would rather
suggest to optimize a unit cell and build the ribbon by translation,
rather than optimizing the whole, in order to avoid numerical errors.
Otherwise you can "add" your contacts to your structure by translating
the contact layer only.

Take care also about the size of your PLs. The carbon-carbon interaction
goes up to the third neighbour, hence the size of the PL should be
larger than the one you are actually using. The error will not be that
dramatic, but usually third neighbour interaction should be taken into
account in graphene. Note also that when you use the "TransportOnly"
option, you don't need a Poisson solver.

Gabriele


>
> so how to I get rid of this error? any kind of help is highly appreciated
>
>  bedamani
>
>
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at dftb-plus.info
> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user


-- 
Dr. Gabriele Penazzi
BCCMS - University of Bremen

http://www.bccms.uni-bremen.de/
http://sites.google.com/site/gabrielepenazzi/

phone: +49 (0) 421 218 62337
fax: +49 (0) 421 218 62770

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.dftb-plus.info/pipermail/dftb-plus-user/attachments/20130117/e839050d/attachment.html>