[Smeagol-discuss] from PHD student in Augsburg Uni.

[Stefano Sanvito]

Dear Mohamed,

This is really too little to understand. What do you mean the charge is not
conserved? Charge is not supposed to be conserved in Smeagol, since the
self-energies introduce a non-hermitian part to the Hamiltonian. In general
however the charge is approximately conserved even at high biases. In
particular if you are dealing with metals, charge should be almost perfectly
conserved.

If you loose a considerable amount of charge (several electrons), usually it
means that the lower bound of the contour integration is set too high. A good
practise is to run a Siesta calculation first and then check where is the
lowest eigenvalue. Then you will set the lower bound of Smeagol integration to
an energy somewhat lower than the lowest caluclated with siesta. Finlly you
will start the Smeagol calcultion using the density matrix already converged
with siesta.

I hope this help,

Stefano

Quoting "Ï.ãÍãÏ ãÕØÝí" <fadlgod at gmail.com>:

> Hello,
> I have a problem in my calculation (24 atoms of Al) I obtained the
> charge conservation at the following values:
> ElectronicTemperature 508 meV
> NEnergReal 1000
> NEnergImCircle 50
> NEnergImLine 3
> NPoles 4
> (if I change any of NEnergImLine, NPoles and ElectronicTemperature
> there is no conservation) . When I calculated I-V there is no
> conservation of charge after V=0 and the number of charge is very very
> different and sometimes it is negative. Also I noticed that at these
> Voltage, in siesta iterations, the energies are positive and the
> change of density matrix is very very large.
> in another case (24 atoms of Cu) I obtained the charge conservation at
> the following values:
> ElectronicTemperature 300 K
> NEnergReal 1000
> NEnergImCircle 500
> NEnergImLine 100
> NPoles 10
> I obtained the same problems. Have you any comment about that
> Best regards
> Mohamed
> --
> Mohamed M. Fadlallah
> Assistant lecturer
> Physics Department
> Benha University
> Benha, Egypt
>
>


--
Prof. Stefano Sanvito
School of Physics and CRANN,
Trinity College Dublin, IRELAND
Tel: +353-1-8963065
www.smeagol.tcd.ie