[DFTB-Plus-User] Different Fermi energy calculated by DFTB+NEGF and only DFTB+
Gabriele Penazzi
gabriele.penazzi at bccms.uni-bremen.de
Sun May 11 13:12:27 CEST 2014
On 05/11/2014 11:47 AM, Enlong Liu wrote:
> Dear Gabriele,
>
> In the attachment is the file I used for calculation.
Hi Enlong, I see some errors in the transport input file.
If you want to consider a bulk Ag electrode, then your structure needs
to be periodic (S flag and lattice vector to be specified).
What you're really calculating here is a sort of Ag nanowire as
electrode, only periodic in transport direction, therefore I don't
wonder that the Fermi level is strange.
Note also that for the cluster case, the k-point sampling would be wrong
(4 should be the sampling along the first lattice vector which is
automatically determined in the cluster case).
Try to define correctly the lattice vectors in x and y direction and set
the structure as periodic. The k-point sampling is ok in this case,
because when the system is periodic the lattice vector in the transport
direction is not automatically determine.
The ShiftAccuracy values are too large. Try to use the defaults, if the
code complains you probably need a larger surface slab. You can set a
smaller ShiftAccuracy value, but not that large (some percent of atomic
units, at most)
I suggest you also to use a small finite temperature (10K for example)
to help convergence. I don't have your sk files, therefore I can check
input consistency but not quantitative results.
>
> Thanks a lot for your help!
>
> Best regards,
>
>
> 2014-05-11 16:50 GMT+08:00 Gabriele Penazzi
> <gabriele.penazzi at bccms.uni-bremen.de
> <mailto:gabriele.penazzi at bccms.uni-bremen.de>>:
>
> On 05/11/2014 10:31 AM, Enlong Liu wrote:
>> Dear Gabriele,
>>
>> I checked my system according to your suggestion in these days,
>> Here is the feedback:
>>
>> 1. My system is metal. Since my system is a nanotube standing
>> vertically on a part of Ag(111) slab and I want to get the
>> transmission function through the contact, one of the contact is
>> part of the Ag slab.
>>
>> 2. You mean that in contact calculation of Ag slab, the k point
>> and weights should be the same as DOS and band structure
>> calculation? But in the latter, the total energy and Fermi energy
>> changes according to the k points. I think maybe convergence
>> should be reached in this case to determine the k point?
>
> Yes, if you compare two calculations the k-point sampling should
> be the same. Sometimes this can be confusing in 1D systems due to
> the way the k-sampling is defined in the transport input file, but
> there should be no confusion for a periodic case.
>
> I am not sure what you mean for band structure calculation, as in
> this case you first calculate the self-consistent charge density
> with a given k-point sampling (the one which matters for getting
> the right solution) and then the band structure with a non-scc
> calculation and a different k-point sampling (the one which
> matters only for fixing the coordinates of your plot in reciprocal
> space).
>
> Also I didn't understand if your Fermi level is consistent respect
> to the DOS. Could you send the input files you use for the two
> calculations?
> For periodic structure, there could be some issue due to the
> reference potential in the poisson solver, which could be solved
> by using the poisson solver for the contact calculation as well.
> Anyway, if you send the input files I can try to verify if the
> issue is in the input or in the code.
>
>
>
>>
>> 3. I am not very clear about this tip, since I don't know where
>> can I get the reference.
>
> I just mean that if you compare the contact calculation and a
> periodic dftb calculation, do it using the same version of the
> code. Fermi level can differ, the important is that they are
> consistent with the DOS and with the rest of the calculation.
>
>>
>> Thanks!
>>
>> Kind regards,
>>
>>
>> 2014-05-09 22:53 GMT+08:00 Gabriele Penazzi
>> <gabriele.penazzi at bccms.uni-bremen.de
>> <mailto:gabriele.penazzi at bccms.uni-bremen.de>>:
>>
>> On 05/08/2014 05:06 PM, Enlong Liu wrote:
>>> Dear all,
>>>
>>> I am doing transport calculation to boron nanotubes (BNT).
>>> But I find out that there is a huge difference in fermi
>>> energy in different cases.
>>>
>>> In the transport calculation, I used DFTB+NEGF to calculate
>>> the contact firstly, which is part of BNT at one end and the
>>> result is -10eV approximately. But if I just calculate the
>>> DOS or band structure to the same BNT part, the fermi energy
>>> is just -4eV. It is interesting since I think all other
>>> parameters are the same, only the calculation situation is
>>> different, one is just DFTB+ for DOS, the other includes
>>> NEGF to calculate contact properties before doing transport
>>> calculation.
>>>
>>> I tried some other system like Si or Ag. There is the same
>>> problem, fermi energy difference in common calculation and
>>> transport calculation. Could any one help me with that?
>>
>> Dear Enlong,
>>
>> we need to be a bit careful here as there is a couple of issues.
>>
>> First: is the system a semiconductor or a metal or a
>> degenerate semiconductor?
>> In the first case, as long as the Fermi level is in the gap,
>> at zero temperature there's no difference between
>> quantitative differences in the numerical value.
>>
>> Second: be careful with the k point sampling, sometimes this
>> can look confusing. Check the manual to verify that your
>> k-points are consistent between dos and band structure
>> calculation.
>>
>> Third: compare consistently. I am not sure whether it makes
>> difference or not, but to be safe just compare quantities you
>> get with the same version of the code as the Fermi level,
>> being an energy, is defined up to a given reference.
>>
>> Fourth: if all the above is verified, in some specific cases
>> there could be still unknown issues. My experience is that in
>> non-periodic systems everything works well. In periodic
>> systems there could be more delicate cases due to the
>> different boundary conditions in the poisson. Nevertheless
>> the huge difference (6 eV) makes me believe that you are in
>> one of the previous cases. It is a good idea that you check
>> your equilibrium calculation and verify that at the interface
>> between contact and device region discontinuities in the
>> potential are not large, because that is often a signature of
>> a wrong fermi level.
>>
>> Regards,
>> Gabriele
>>
>>
>>>
>>> Thanks a lot!
>>>
>>> Best regards,
>>> Enlong
>>>
>>> --
>>> Faculty of Engineering at K.U <mailto:Engineering at K.U>. Leuven
>>> BIOTECH at TU Dresden
>>> Email:liuenlong20 at gmail.com <mailto:liuenlong20 at gmail.com>;
>>> enlong.liu at student.kuleuven.be
>>> <mailto:enlong.liu at student.kuleuven.be>;
>>> enlong.liu at biotech.tu-dresden.de
>>> <mailto:enlong.liu at biotech.tu-dresden.de>
>>> Mobile Phone: +4917666191322 <tel:%2B4917666191322>
>>> Mailing Address: Zi. 0108R, Budapester Straße 24, 01069,
>>> Dresden, Germany
>>>
>>>
>>> _______________________________________________
>>> DFTB-Plus-User mailing list
>>> DFTB-Plus-User at dftb-plus.info <mailto: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 <tel:%2B49%20%280%29%20421%20218%2062337>
>> fax:+49 (0) 421 218 62770 <tel:%2B49%20%280%29%20421%20218%2062770>
>>
>>
>> _______________________________________________
>> DFTB-Plus-User mailing list
>> DFTB-Plus-User at dftb-plus.info
>> <mailto:DFTB-Plus-User at dftb-plus.info>
>> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
>>
>>
>>
>>
>> --
>> Faculty of Engineering at K.U <mailto:Engineering at K.U>. Leuven
>> BIOTECH at TU Dresden
>> Email:liuenlong20 at gmail.com <mailto:liuenlong20 at gmail.com>;
>> enlong.liu at student.kuleuven.be
>> <mailto:enlong.liu at student.kuleuven.be>;
>> enlong.liu at biotech.tu-dresden.de
>> <mailto:enlong.liu at biotech.tu-dresden.de>
>> Mobile Phone: +4917666191322 <tel:%2B4917666191322>
>> Mailing Address: Zi. 0108R, Budapester Straße 24, 01069, Dresden,
>> Germany
>>
>>
>> _______________________________________________
>> DFTB-Plus-User mailing list
>> DFTB-Plus-User at dftb-plus.info <mailto: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 <tel:%2B49%20%280%29%20421%20218%2062337>
> fax:+49 (0) 421 218 62770 <tel:%2B49%20%280%29%20421%20218%2062770>
>
>
> _______________________________________________
> DFTB-Plus-User mailing list
> DFTB-Plus-User at dftb-plus.info <mailto:DFTB-Plus-User at dftb-plus.info>
> http://www.dftb-plus.info/mailman/listinfo/dftb-plus-user
>
>
>
>
> --
> Faculty of Engineering at K.U. Leuven
> BIOTECH at TU Dresden
> Email:liuenlong20 at gmail.com <mailto:liuenlong20 at gmail.com>;
> enlong.liu at student.kuleuven.be
> <mailto:enlong.liu at student.kuleuven.be>;
> enlong.liu at biotech.tu-dresden.de <mailto:enlong.liu at biotech.tu-dresden.de>
> Mobile Phone: +4917666191322
> Mailing Address: Zi. 0108R, Budapester Straße 24, 01069, Dresden, Germany
>
>
> _______________________________________________
> 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://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20140511/47f5b5cd/attachment.htm>
More information about the DFTB-Plus-User
mailing list