Nadia you are not right. A denser k-point sampling COULD be used, but it is not necessary. In the 2-dimensional case you have the E(k_x,k_y,k_z=0) function for the energy, instead of the more general 3-dimansional E(k_x,k_y,k_z). Therefore, the accuracy of the E(k) expression would be the same. Just ensure that if you have a denser k-points sampling, DOS converges.
Best Zacharias Fthenakis > Dear Dr. Papior, > > Since my system is a slab, so the scattering occurs in two dimensions (the > transport direction and also the transverse direction), According to two > directions, I think that denser k point sampling along two directions must > be used to calculate the transmission. Am I right? > > Thanks in advance, > > Best regards, > Nadia Salami > > On Sun, Jul 26, 2015 at 11:53 AM, Nadia Salami <[email protected]> > wrote: > >> Dear Dr. Papior, >> >> Since my system is a slab, so the scattering occurs in two dimensions >> (the >> transport direction and also the transverse direction), According to two >> directions, I think that denser k point sampling along two directions >> must >> be used. Am I right? >> >> Thanks in advance, >> >> Best regards, >> >> Nadia Salami >> >> On Sat, Jul 25, 2015 at 11:46 AM, Nick Papior <[email protected]> >> wrote: >> >>> >>> >>> 2015-07-25 6:55 GMT+00:00 Nadia Salami <[email protected]>: >>> >>>> Dear Transiesta users, >>>> >>>> I don’t know how to determine correct k point sampling for the >>>> Transiesta as well as Tbtrans calculations. >>>> >>> You converge using the same principles as standard DFT, for transport >>> calculations however, you converge the averaged transmission. >>> >>>> In details, I have determined k-point sampling for the electrode of my >>>> system from the energy convergence test, that leads to the following >>>> sampling >>>> >>>> %block kgrid_Monkhorst_Pack >>>> >>>> 1 0 0 0.0 >>>> >>>> 0 2 0 0.0 >>>> >>>> 0 0 4 0.0 >>>> >>>> %endblock kgrid_Monkhorst_Pack >>>> >>>> >>>> >>>> Must the k grid along the transport direction (i.e. kz) be increased >>>> to >>>> simulate the semi-infinite lead (electrode)? >>>> >>> I would always use as many k-points in the z-direction as you can (the >>> more k-points, the better description of the self-energy you get). >>> I typically use 100, if that proves too much, I try with 50. >>> >>>> Moreover I know that, the denser k-point sampling must be used to >>>> calculate the transmission using Tbtrans utility. Must kx and ky be >>>> increased? For Tbtrans calculation or both of Transiesta and Tbtrans >>>> calculations? (Also I know that kz only is utilized to calculate >>>> siesta >>>> calculation in the scattering region.) >>>> >>> Transiesta calculates the density, tbtrans calculates the transport. >>> >>>> Finally, how to converge the k-point sampling for the transmission >>>> calculation? >>>> >>>> It will be highly appreciated your comments and guidances. >>>> >>>> Thanks in advance, >>>> >>>> Best regards, >>>> >>>> Nadia Salami >>>> >>>> >>>> >>> >>> >>> >>> -- >>> Kind regards Nick >>> >> >> > ********************************************************* Dr Zacharias G. Fthenakis Research Associate Institute of Electronic Structure and Laser (I.E.S.L.) Foundation for Research and Technology Hellas (FO.R.T.H.) Vassilika Vouton P.O. Box 1527 71003 Heraklion Crete Greece Phone +30 2810 391824 FAX +30 2810 391305 webpage: http://esperia.iesl.forth.gr/~fthenak **********************************************************
