On Sat, Jan 11, 2020 at 11:00 AM Victor Bermudez <[email protected]> wrote:
Recent versions of QE print a table of "atom charges" upon convergence of > each SCF cycle. What do these numbers mean ? they are meant to give a vague idea of where the magnetization is located, I think. They are electronic charges (number of electrons, I guess) and magnetization - difference between up and down charges, integrated in a spherical region of radius r_m (printed in output). Paolo I'm aware that accurate estimates of atom charges require something like a > Bader analysis, but I'd still like to know how to interpret these "atom > charge" values. > > For example, here's output (input given below) for a free NO2 > molecule. This is just a "simple" example of a general result. > > Magnetic moment per site: > atom: 1 charge: 1.9288 magn: 0.1688 constr: > 0.0000 > atom: 2 charge: 2.9353 magn: 0.1673 constr: > 0.0000 > atom: 3 charge: 2.9339 magn: 0.1665 constr: > 0.0000 > > Atom 1 is N, and atoms 2 and 3 are O. The core charges (Zval) are +5 for N > and +6 for O, giving a total of +17. For a charge-neutral molecule I would > therefore expect the "charge" number to sum to 17 if these numbers > represent electron densities. Instead the sum is 7.798 |e|. If the charges > are in Rydberg atomic units (1 au = |e|/rt(2)) then the sum is actually > 5.515 |e|. > > Another question I have concerns the "magnetization" values. What > quantities do they represent, and what are the units ? > > Here's the input (for a cubic lattice of molecules with a > 20-Angstrom lattice constant): > > ******************************************************************************* > &CONTROL > calculation='relax', > restart_mode='from_scratch', > title='free NO2; start with #237406 opt.; 200/800 cut-offs', > pseudo_dir="/p/home/bermudvm/Pseudo", > outdir="/p/work1/workspace/bermudvm/2753418.THUNDER/tmp", > ! outdir='/lustre/cmf/scratch/b/bermudez/171133/', > etot_conv_thr=1.0D-5, > forc_conv_thr=1.0D-4, > nstep=500, > verbosity='default', > max_seconds=21500.0 > / > > &SYSTEM > ibrav=1, > a=20.0, > nat=3, > ntyp=2, > ecutwfc=200.0, > ecutrho=800.0, > occupations='smearing', > smearing='gaussian', > degauss=0.001, > nspin=2, > starting_magnetization(1)=0.20, > starting_magnetization(2)=0.0, > ! nosym=.true., > ! nbnd=14, > vdw_corr='TS' > / > > &ELECTRONS > electron_maxstep=100, > conv_thr=1.0D-9, > mixing_beta=0.7, > mixing_mode='plain' > ! mixing_ndim=8 > / > > &IONS > ion_dynamics='bfgs', > upscale=10000.D0 > / > > ATOMIC_SPECIES > N 14.0067 N_ONCV_PBE-1.0.upf > O 15.9994 O_ONCV_PBE-1.0.upf > > ATOMIC_POSITIONS angstrom > N -0.016847220 -0.000000025 -0.014422960 > O 1.092527105 0.000000576 0.456691560 > O -1.125679885 -0.000000551 0.457731400 > > K_POINTS gamma > > ************************************************************* > > Any advice would be much appreciated. > > Vic Bermudez > US Naval Research Lab. (retired) > _______________________________________________ > Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) > users mailing list [email protected] > https://lists.quantum-espresso.org/mailman/listinfo/users > -- Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche, Univ. Udine, via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222
_______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
