There sure is. LDA is known to predict qualitatively incorrect magnetic properties for a number of systems, perhaps the most famous case is bulk iron. GGA seems to perform generally better, but you should still treat your results with extreme caution.
2007/12/22, Roberto Sapiens <[EMAIL PROTECTED]>: > > Hi everybody: > > I'm carrying out spin polarized calculations of H-terminated zigzag > graphene ribbons. If the calculations are done with GGA-PBE, I obtain > an absolute spin polarization of carbons on the edges of 0.26. With > the same input, but LDA-CA, I obtain 0.16. I'm calculation > magnetization of a single atom by integrating PDOS_up-PDOS_down up to > Fermi energy. Is there any good reason for that difference? > > []s, > > Roberto > > -- > "If I have seen farther than others it is because of a myopia surgery." >
