On 05/03/2011 05:54 PM, Izaak Williamson wrote: > Dear all, > > I am trying to calculate the cohesive energy of the N2 molecule using the > attached input file (relax.in <http://relax.in/>) and am getting a value of > -16.57 eV. I use DFT with GGA and the pseudo-potential N.pbe-rrkjus.UPF. Other > work [Fuchs et. al., Phys. Rev. B 65, (2002)245212] has performed similar > calculations and obtained values ~10.5 eV. They even list an experimental > value > of 9.76 eV. Why is my value so much higher? Is there anything in my input file > that could be giving inaccurate results? Is it my pseudo-potential that is > causing this problem? > > Thanks for any help. Dear Izaak, how much do you get for an isolated N atom, spin polarized, same cell, same pseudo, same cutoffs? In an old output file I've got: -39.7039435178 Ry for N2 at equilibrium and -19.48996768 Ry for the N atom. Therefore: -0.724 Ry = -9.85 eV (N.pbe-rrkjus.UPF)
Davide
