Sorry, I tend to disagree with Ney Henrique that the things are that bad; please correct me if I am wrong: If basis function of your fragments do not overlap (i.e., the "molecule" regime is recognized) - then, I thought, the calculation is as good as it would be for isolated molecule in the infinite cell. I think, notably an issue of a charged molecule do not pose a problem and do not produce any infinite replicas, because as the "molecule" regime is recognized, the Madelung terms are switched off. (Note that Siesta allows charged systems for cubic cells only, for which she knows how to calculate Madelung sums). I am not so sure about whether there are differences in less sensitive issues, like construction of the average potential, higher-order multipoles, or treatment of empty space (no basis) on a grid... Anything else?
Best regards Andrei Postnikov > Hi all > > @ Gregorio: Yes, you are right! But strictly speaking that is not the same > as running an isolated system. You will have infinite replicas of your > isolated molecule anyway ... and it is problematic if you have charged > species (there will be a rigid potential shift due to the periodic > replication of the charges) although the Markov-Paine correction is > implemented. > > @ Robert : Take a closer look to the Manual ... all this stuff is > addressed > there! > > Cheers > > NH > > 2011/10/6 Gregorio García Moreno <[email protected]> > >> ** >> Yes >> Using supercell aproximation. i.e. Put you cluster or your aisolated >> molecule, within of a very big cells. Thus, there are not interactions >> between molecules of different cells, like in a isolate molecule >> approximation. >> >> El 10/6/2011 2:22 PM, [email protected] escribió: >> >> Dear Users, >> >> I have one probably very basic question. Is it possible to run >> calculation >> (for some cluster system) without periodic boundary conditions in Siesta >> ? >> >> Thanks >> >> Robert Sedlak >> >> >> >> >> > > > -- > Dr. Ney Henrique Moreira > Bremer Center for Computational Materials Science > Am Fallturm 1 > 28359 Bremen > Deutschland > mobile: 0049176-20485882 >
