[CPMD-list] free isolated atom energiesB

[Eduardo Ariel Menendez P]

Hello again,
First of all, thanks to AK, AS, and OY for your comments. Then a few more
calculations and questions.

>
> stupid question: did you already try the equivalent pseudopotentials
> from the goedecker hartwigsen library? at least for LDA there is a very
> large collection, including Te and two version of Cd potentials (normal
> and semi-core). also the way they are defined seems to make them much
> more suitable for those elements than the troullier-martins
> pseudopotentials in separable form. at least this is my (limited)
> experience so far.

 I have used them (Goedecker's PPs), but I find that the cutoff required
for convergence are larger than for MT pseudos.  For O I find the same
behavior in the
atomic calculation. However, see below.

>
>
> USPPs and LANCZOS are not compatible. see the recent
> mail by juerg hutter in the mailing list archives.

Let's turn to Norm conserving pseudopotentials.
>
> finally, i am not so sure, if you can compare the
> absolute values, since you are using different grids
> for the atomic and the cpmd calculation.

Naive thought: I guess I should obtain comparable results, if one wishes
to trust in CPMD
molecular dynamics. ?What about the forces? At least, the Born-Oppenheimer
MD is likely to give wrong forces?

Ari Seitsonen pointed that
" 1) Like you noticed, the occupation numbers 2, 2, 1, 1 do not make really
   sense; well, this is leads to a non-spherical atom (p_x, p_y and p_z
   are occupied differently), which is used in the pseudo potential
   program. You should use an equal occupation of the p levels if you want
   to compare with Paolo Giannozzi's program (hmm, keyword 'OCCUPATION'?)"

Then I procceeded to calculate the O atom (with Goedecker's PP) keeping
the 3 2p orbitals with
a population of 4/3. I found convergence with the following sequence:
1)
&CPMD
  OPTIMIZE WAVEFUNCTION
  MAXSTEP
  150
  CENTER MOLECULE ON
&END

&DFT
  NEWCODE
  FUNCTIONAL LDA
  LDA CORRELATION PZ
  SLATER
    0.666667
  GC-CUTOFF
  0.1E-07
&END

&SYSTEM
  SYMMETRY
  0
  POISSON SOLVER TUCKERMAN
  ANGSTROM
  CELL
    10.0  1.0  1.0  0  0  0
  CUTOFF SPHERICAL
   155.0
  STATES
   4
  OCCUPATION FIXED
   2.0 1.33333333333333 1.33333333333333 1.33333333333333
&END

&ATOMS
*O-q6
   LMAX=P LOC=P
   1
    0.000000    0.000000    0.000000
&END


2) Using the RESTART file, I continued the optimization with
Lanczos diagon, and fixing the occupation to keep the atom spherical
Changed
&CPMD
  OPTIMIZE WAVEFUNCTION
  RESTART DENSITY
  LANCZOS DIAGONALISATION ALL
  MAXSTEP
  150
  CENTER MOLECULE ON
&END
and in section SYSTEM I set
  STATES
   6
  OCCUPATION FIXED
   2.0 1.33333333333333 1.33333333333333 1.33333333333333 0.0 0.0
&
With the above, I obtained eigenvales not converged.

3) Then I made a calculation of te eigenvalues
&CPMD
  RESTART DENSITY WAVEFUNCTION
  KOHN-SHAM ENERGIES
   0
  LANCZOS DIAGONALISATION ALL
  MAXSTEP
  150
  CENTER MOLECULE ON
&END

  STATES
   6
  OCCUPATION FIXED
   2.0 1.33333333333333 1.33333333333333 1.33333333333333 0.0 0.0


Then, I got the eigenvalues:
 EIGENVALUES(EV) AND OCCUPATION:
     1    -23.6274394     2.000        2     -9.0449296     1.333
     3     -9.0448316     1.333        4     -9.0448264     1.333
     5     -0.2474729NC   0.000        6      1.2166765NC   0.000
 CHEMICAL POTENTIAL =                             1.2166765461 EV

Eigenvalues 5 and 6 are not converged yet, but at least there are 3 almost
equal energies close to the reference 2p energy, as well as the
 AM> 2s     -0.871362 Ha= -23.70104 eV
> AM> 2p     -0.338381 Ha = -9.20 eV

There are still differences with the reference energies that are larger
than the standard criterium of 0.01 ev/atom.
Are this errors not important for the calculation of forces?

Is there any other simple test for the new generated pseudopotentials
after being translated to the CPMD format?


Best regards,
Eduardo