ELF3D and RHO3D

[SangHwan Kim]

Dear all,

When elf and rho calculated, I found many negative values in both files.

Please look at my input file and let me know what I made a mistake.

tasks
   1
  10
  20

avec
   17.73205620       0.000000000       0.000000000
  -8.866028100       15.35641113       0.000000000
   0.000000000       0.000000000       9.672185268

atoms
   4                                    : nspecies
'Li.in'                                 : spfname
   2                                    : natoms; atposl, bfcmt below
    0.66666667    0.33333333    0.27500000    0.00000000  0.00000000  0.00000000
    0.33333333    0.66666667    0.77500000    0.00000000  0.00000000  0.00000000
'Ti.in'                                 : spfname
   1                                    : natoms; atposl, bfcmt below
    0.00000000    0.00000000    0.83670000    0.00000000  0.00000000  0.00000000
'I.in'                                  : spfname
   6                                    : natoms; atposl, bfcmt below
    0.67844000    0.64855000    0.71140000    0.00000000  0.00000000  0.00000000
    0.35145000    0.02989000    0.71140000    0.00000000  0.00000000  0.00000000
    0.97011000    0.32156000    0.71140000    0.00000000  0.00000000  0.00000000
    0.02989000    0.67844000    0.21140000    0.00000000  0.00000000  0.00000000
    0.32156000    0.35145000    0.21140000    0.00000000  0.00000000  0.00000000
    0.64855000    0.97011000    0.21140000    0.00000000  0.00000000  0.00000000
'O.in'                                  : spfname
  18                                    : natoms; atposl, bfcmt below
    0.81090000    0.86200000    0.58480000    0.00000000  0.00000000  0.00000000
    0.13800000    0.94890000    0.58480000    0.00000000  0.00000000  0.00000000
    0.05110000    0.18910000    0.58480000    0.00000000  0.00000000  0.00000000
    0.55080000    0.79460000    0.00130000    0.00000000  0.00000000  0.00000000
    0.20540000    0.75620000    0.00130000    0.00000000  0.00000000  0.00000000
    0.24380000    0.44920000    0.00130000    0.00000000  0.00000000  0.00000000
    0.49580000    0.60930000    0.53820000    0.00000000  0.00000000  0.00000000
    0.39070000    0.88650000    0.53820000    0.00000000  0.00000000  0.00000000
    0.11350000    0.50420000    0.53820000    0.00000000  0.00000000  0.00000000
    0.94890000    0.81090000    0.08480000    0.00000000  0.00000000  0.00000000
    0.18910000    0.13800000    0.08480000    0.00000000  0.00000000  0.00000000
    0.86200000    0.05110000    0.08480000    0.00000000  0.00000000  0.00000000
    0.75620000    0.55080000    0.50130000    0.00000000  0.00000000  0.00000000
    0.44920000    0.20540000    0.50130000    0.00000000  0.00000000  0.00000000
    0.79460000    0.24380000    0.50130000    0.00000000  0.00000000  0.00000000
    0.88650000    0.49580000    0.03820000    0.00000000  0.00000000  0.00000000
    0.50420000    0.39070000    0.03820000    0.00000000  0.00000000  0.00000000

    0.60930000    0.11350000    0.03820000    0.00000000  0.00000000  0.00000000

vkloff
   0.0, 0.0, 0.0

xctype
  2

maxscl
  1

autormt
  .true.

nempty
50

beta0
  0.2

scissor
0.0367

sppath
  '../../../software/exciting/species/'

rgkmax
6

ngridk
  2 2 3

plot3d
1 1 1     : nup3d
101 101 55  : np3d

Best Wishes,
Sang-Hwan (Postdoc)
University of Houston

[exciting]

Hi Sang-Hwan,

You may occasionally get ELF values which go slightly below zero, and sometimes even the density may be slightly negative, although this is unusual.

Are all the other output quantities (total energy, eigenvalues, occupancies, bandstructure, DOS, etc.) reasonable?

Cheers,
Kay.

[SangHwan Kim]

Dear Kay,

I plotted TDOS and PDOSs.
And they are reasonable.
Also, their charge is reasonable, too.

Can it be related to small grid points?

Best,
Sang-Hwan

[Anton]

Hi Sang-Hwan!
The charge density could be negative due to numerical algorithms. Try to plot RHO with the following switches:

lmaxvr
  8
lmaxinr
  8

and you will get the smooth positive charge-denisty.

Cheers,
Anton.