[Pw_forum] Problem with PAW vs. USP

[Vic Bermudez]

Hello Lorenzo,

	Thank you for your comments. Yes, getting the free-atom energy right is
tricky. Below is a sample input file that, when used with a USPP and not a
PAW PP, gives the correct Zr free-atom energy. I say "correct" because the
correct cohesive energy (i.e., very close to experiment) is then obtained
for the bulk metal. The starting magnetization value comes from the fact
that 12 electrons (2 of which are unpaired in the (3)F ground state) are
treated explicitly, whence starting_magnetization(1) = 2/12 = 0.16667.
	This same input file does not work correctly in a PAW calculation, probably
because of the use of the OCCUPATIONS card. For PAW I was able to get the
correct free-atom energy by removing:

	occupations='from_input',
	one_atom_occupations=.TRUE.,
	starting_magnetization(1)=0.16667

and

	OCCUPATIONS
	1 1 1 1 1 1 0 0 0 1 0 0 0
	1 1 1 1 0 0 0 0 0 1 0 0 0

and simply using tot_magnetization=2.0. The resulting free-atom energy is
then very close to that given in Giovanni Cantele's message from earlier
today, and the correct cohesive energy is then obtained.
	Hopefully other users will find this helpful.

Best Wishes,
Vic Bermudez

Victor M. Bermudez
E-mail: bermudez at alum.mit.edu


________________ Sample Free-Atom Input File ____________


&CONTROL
calculation='scf',
title='free Zr atom - should give (3)F ground state - using PBE/PAW',
pseudo_dir='/lustre/cmf/scratch/b/bermudez/QE_PP/',
verbosity='high',
wf_collect=.TRUE.
/

&SYSTEM
ibrav=1,
a=40.0,
nat=1,
ntyp=1,
nbnd=13,
ecutwfc=40.0,
ecutrho=320.0,
nosym=.TRUE.,
nspin=2,
occupations='from_input',
one_atom_occupations=.TRUE.,
starting_magnetization(1)=0.16667
/

&ELECTRONS
electron_maxstep=200,
conv_thr=1.0D-7,
mixing_beta=0.7D0,
mixing_ndim=8,
mixing_mode='plain',
startingwfc='random'
/

ATOMIC_SPECIES
Zr 91.22   Zr.pbe-spn-kjpaw_psl.0.2.3.UPF {** NOTE ** paw should not be used
here !}

ATOMIC_POSITIONS angstrom
Zr   0.000 0.000 0.000

K_POINTS automatic
1 1 1 0 0 0

OCCUPATIONS
1 1 1 1 1 1 0 0 0 1 0 0 0
1 1 1 1 0 0 0 0 0 1 0 0 0