DFT calculations for two probe systems

[jerry]

Dear,
    Recently, i have read some papers that attribute the discrepancy between calculated conductance and measured one to the underestimation of the HOMO-LUMO gap of DFT methods. Usually, this discrepancy is orders of magnitude. I want to know whether this kind of argument is reasonale, as the main effect of the underestimation of HOMO-LUMO gap is the decrease of the conductance gap.

[Nordland]

It is hard to say anything in general, but what kind of systems are you thinking about ? A Molecular junction?

[jerry]

Yes, a molecular junction with thiol ends. For instance, DBT.

[jerry]

Can anyone help me? Thanks!

[kstokbro]

The homo-lumo gap is one of the most important parameters, however, it is also important how the metal fermi level aligns with the molecular levels.
There are also other important aspects, forinstance what is the width of the resonance. Thus, this is a very complex question without a definite answer.

I suggest you consult  the following papers:

Role of the virtual orbitals and HOMO-LUMO gap in mean-field approximations to the conductance of molecular junctions
A. Cehovin, H. Mera, J. H. Jensen, K. Stokbro, and T. B. Pedersen
Phys. Rev. B 77, 195432 (2008) – Published May 22, 2008

Are Kohn-Sham Conductances Accurate?
H. Mera and Y. M. Niquet
Phys. Rev. Lett. 105, 216408 (2010) – Published November 19, 2010

and:
42) Self-consistent GW calculations of electronic transport in thiol- and amine-linked molecular junctions
M. Strange, C. Rostgaard, H. Hakkinen, and K. S. Thygesen
Phys. Rev. B 83, 115108 (2011)
http://dcwww.camd.dtu.dk/~thygesen/my_papers_pdf/GW_BDT_BDA.pdf

[jerry]

Thank you very much! By the way, can you tell me why the mulliken population of an atom in a molecule is not integrals? Does the temperature effect on the Fermi function leads to this? Because the limit of integration is the whole space.

[kstokbro]

The mulliken population is a projection of the total number of electrons onto individual atoms and can take any value. The only requirement is that the sum of the mulliken populations give the total charge.

[jerry]

Thanks a lot!
When i calculated the molecular junction using ATK according to the manual, i found that the Fermi level of the electrodes is just between the HOMO and LUMO of this molecule. Does this mean that only fractional mounts of electrons transfer between the molecule and electrodes to keep equilibrium When the applied bias is 0?