how to choose the R e al a xis p oin t d e n sit y?

[njuxyh]

Hi all:

I read the following tutorial:   http://docs.quantumwise.com/_downloads/atk_transport_calculations.pdf

NEGF contour integral NEGF contour integral For finite-bias calculations, a potentially important convergence parameter is the R e al a xis p oin t d e n sit y , which is specified under C o n t o u r in t e g r al p a r a m e t e r s . This parameter determines how many energy points are used when integrating the left and right spectral density matrices along the real energy axis between the chemical potentials of the left and right electrodes. For the present calculation you can proceed with the default value, but be aware that a convergence test of this parameter should in general be performed.


so, if the self-consistent calculation of device under bias did not converge?  should I increase the real axis point density to see the results? 

[Jess Wellendorff]

Poor SCF converge can have a number of reasons, and increasing the point densities for the NEGF contour integral may or may not help you. I would first try out slower density mixing and make sure to restart the calculation from a converged calculation at similar bias.

[Daniele Stradi]

Dear njuxyh,

That is one of the important parameters that determines the accuracy of a finite bias calculation, but normally the default is quite safe (assuming that you are using FHI pseudopotentials). If the SCF cycle of your NEGF calculation gets very close to convergence, e.g. the variation in the Hamiltonian oscillates around 10^-3 eV (the default threshold for convergence is 10^-4), increasing the number of energy points can help.

On the other hand, if your SCF cycle blows up completely, it normally means that there is something else which goes wrong. For example, the density matrix used as an initial starting guess is not good enough, or that the bias you are applying is too large for your system. Of course, these are just guesses, it is difficult to say more without seeing the actual input.

Feel free to attach the input here or send it by mail if you need further assistance.

Regards,
Daniele.