Dear Masoudeh -indeed, it looks like a big confusion. A pity because these issues were so many timesaddressed in tutorials and - I am sure - in the mailing list.OK, slowly: > I can’t optimize three parameters such as k point, mesh cut off and lattice constant. These "parameters" are not in the same standing. The lattice constant is your result,you can indeed optimise it by varying it and looking for total energy minimum. The mesh cutoff and k-points a responsible for the accuracy of calculation.There are no "optimal" value of them that would minimise the total energy.Typically, as you increase them the total energy will consistently go down.You just make a free choice of these parameters driven by analysis of whatis the accuracy of the total energy you really need. > Should I do all these process in doped tinoxide again? If the given mesh > cutoff value is OK for pure SnO2 -and- also for
----- masoudeh maleki <[email protected]> a écrit : >Dear Andrei,Thank you so much for your email. I'm so confused because I'm >trying to dope tinoxide with nitrogen by siesta code and I can’t optimize >three parameters such as k point, mesh cut off and lattice constant. I mean >that plot of the total energy against these parameters isn’t as well as pure >tinoxide. Before doping I optimize these parameters in bulk tinoxide and then >relaxed and after relaxed the structure. Should I do all these process in >doped tinoxide again? And would you please kindly guide me what should I do to >have acceptable plots in doped case? Should I change the coordinates doped >nitrogen atom manually or subsituating nitrogen instead of oxigen or tin is >enough ? Yours SincerelyMasoudeh Maleki On Wed, Sep 27, 2017 at 11:31 PM Andrei Postnikov <[email protected]> wrote: ----- masoudeh maleki <[email protected]> a écrit : >Dear sir, I'm using siesta code to investigate the effect of doping on bandstructure and PDOS of periodic crystal of tin oxide with tetragonal unitcell. I made a 2*2*2 supercell of it, and replaced one O atom with my dopant. According to the siesta manual, can l use NetCharge and simulatedoping together for my bulk structure to simulate doping? Or it is impossible? And if it is incorrect how can I dope my bulk structure in siesta? Yours sincerelyMasoudeh maleki Dear Masoudeh, if you explicitly add the dopant atom to your system, you don't need any NetChargesince your system is not charged. (The dopant atom comes with its correct nuclear chargeand number of electrons). As explained in the manual, the NetCharge option is providedto treat charged finite systems (molecules). Best regards Andrei Postnikov -- -- prof. Andrei Postnikov -- tel. +33-372749149 -- cell +33-636565128 -- University of Lorraine - Laboratoire de Chimie/Physique - A2MC ICPM, 1 Bd Arago - BP 95823, F-57078 Metz Cedex 03, France ------------------------------------------------------------------------
