Hi, I am new to ELK code. I was trying to study a transition metal atom adsorbed on a crystal surface. To do this, I need to know the energy, and probably the charge density, of the atom. Can someone tell me what are the appropriate values of the parameters for transition metal atoms such as Cu? I tried to use larger box size (more than 10*10*10 angstroms), rgkmax, gmaxvr, but the calculations all led to unstable energies (very big oscillations). The only stable energy I got so far is from a box size 9*9*9 with default rgkmax and gmaxvr values. But I am not sure the energy is converged or not.
This is the good reference for atomic energies:
NIST Atomic Reference Data for Electronic Structure Calculations:
Thanks for the reference. I saw the energy from my calculations is in the ball park of that from the reference. I still need to make sure the energy from my calculations is converged.
You need to be very cautious about many transition metal atoms in DFT:
Most programs don't provide such fine grained control over orbital occupations and this is the reason why many, let's say cohesive energies, are (or should be) never published.
Cu is fortunately not so extreme, try to reproduce atomization energy of Cu2 from doi/10.1063/1.2162161 (note that these are non-relativistic results) - if you get it within 0.1 eV your Cu atom is probably on a good way.
Check also this thread about calculations of atoms/molecules: https://sourceforge.net/projects/elk/forums/forum/897820/topic/5163717
Thanks a lot for the information. It is very useful to me.
Log in to post a comment.