ABINIT Discussion Forums

Hi all,

When performing total energy and density of states calculations for isolated atoms using PAW datasets, can I know what is the optimal box size (acell) to use? According to tutorial 1, acell of 10 Bohr radius is used for hydrogen, but it seems to take a different value for a different element in another tutorial. I also found that when performing a spin-polarized calculation with an isolated atom, choosing an optimum box size yields electron orbital energy levels that are degenerate for the spin up and down cases, while for all other sizes the spin up and down energy levels are not degenerate. Does this mean that the optimum box size to use corresponds to the case where the spin up and down cases are degenerate in energy, for a non-magnetic material?

Thanks a lot!
Yesheng

You should simply do a convergence study on the input variable acell for your isolated atom. In my experience, 10 would typically not be large enough, should be something more like 20 or higher.

Hi Prof. Zwanziger,

Thank you very much for the note. I appreciate it.