[DFTB-Plus-User] Carbon nanotube in the dftb+ program

[Gabriele Penazzi]

Hi George,

it's a problem with the precision of the gen file. In the error message 
you get a list of errors for each atom, which are quite small but above 
the tolerance of the code. The tolerance is by default very low (1e-8 
a.u.), you can override this setting by adding the line

ShiftAccuracy = 1e-5

in the block contact. With the new version of the code the default 
tolerance is lower (1e-5), which probably makes sense because structures 
are provided as formatted input files with a somewhat limited precision.

Best,
Gabriele


On 11/24/2015 03:21 PM, George D Kubas wrote:
>
> Hello,
>
>
> I aplogize. I sent the wrong hsd file. My correct file has the 
> geometry section as such.
>
> Transport {
>   Geometry {
>     Device {
>       AtomRange =1 160
>     }
>     Contact {
>       Id = "source"
>       AtomRange =161 288
>     }
>     Contact {
>       Id = "drain"
>       AtomRange =289 416
>     }
>
>   }
>
>
> The same error persists with this code. I will check out the fortran 
> utility you attached to see if this resolves my problem. Thank you for 
> your help.
>
>
>
>
>
> ------------------------------------------------------------------------
> *From:* dftb-plus-user-bounces at mailman.zfn.uni-bremen.de 
> <dftb-plus-user-bounces at mailman.zfn.uni-bremen.de> on behalf of 
> Gabriele Penazzi <penazzi at uni-bremen.de>
> *Sent:* Monday, November 23, 2015 10:08 PM
> *To:* dftb-plus-user at mailman.zfn.uni-bremen.de
> *Subject:* Re: [DFTB-Plus-User] Carbon nanotube in the dftb+ program
> Hi George,
>
> there is some error in the setup. The unit cell of the CNT is 32 
> atoms, the contacts must be two identical layers. In your input file 
> the contact has 160 atoms, corresponding to 5 unit cells. Clearly the 
> two layers can not be two rigidly shifted layers because they'd need 
> an even number of unit cells.
>
> The code write out the lattice vector extracted for the contact when 
> there's an error, you see there's some problem because it has a non 
> zero component in the x direction.
>
> There's a tiny fortran utility I sent in mailing list 
> (https://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/2015/002008.html) 
> which may help you, it is used to repeat a principal layer N times in 
> a way which satisfies the ordering, see also the online tutorial.
>
> Last, if you can compile it consider to use the MPI version, the input 
> format has changed a bit and the MPI format contains the latest one 
> which we'll try to support now on.
>
> Best,
> Gabriele
>
>
> On 11/23/2015 11:42 PM, George D Kubas wrote:
>>
>> Hello,
>>
>>
>> I am quite new to the program and am trying to set up an 8,0 cnt 
>> geometry in a .gen file. When I run the dftb+ program on the source 
>> contact file I keep getting the error "Contact 1 does not consist of 
>> two rigidly shifted layers". I used 
>> http://turin.nss.udel.edu/research/tubegenonline.htmlto generate the 
>> atomic positions in a Gaussian format and 13 cell units long. I 
>> thought I set up the correct order of device, source layer closest to 
>> device, source layer furthest, drain layer closest to the device, and 
>> drain layer furthest. I assume my error lies in the self created .gen 
>> file. Any help would be greatly appreciated. Thank you for your time.
>>
>>
>> George
>>
>>
>> _______________________________________________
>> DFTB-Plus-User mailing list
>> DFTB-Plus-User at mailman.zfn.uni-bremen.de
>> https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user
>
>
>
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at mailman.zfn.uni-bremen.de
> https://mailman.zfn.uni-bremen.de/cgi-bin/mailman/listinfo/dftb-plus-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20151125/6a849ed9/attachment.html>