As Sanjay and Andrei have pointed out the number of core
shells seems to be wrong. I attach the output of the Fritz Haber
Institute code with the input obtained from Abinit's website.
This result leads to ghost states in siesta, but you can use
the radii as an initial guess.
The radii (for the s, p, d and f) are: 3.04 4.04 3.04 3.04 Bohr
They also include a core charge density (for the calculation
of the nonlinear core corecctions) with a radius of 1.00 Bohr
Best regards,
Eduardo
#info /Users/edu/bin/psgen-old ncpp 16:08:28 Apr 08 2008 edu
fhi pseudopotential tool fhipp - version rev 1998
chemical symbol no
nuclear charge 92.00
total charge 0.00
number of core states 15
number of valence states 3
exchange-correlation model 8 LDA CA Perdew/Wang 1991
scalar-relativistic mode
parameters radial mesh 573 1.024700 0.679348E-04
=== all-electron atom ===
< n l occupation eigenvalue(eV)
< 1 1 0 2.0000 -116025.5812
< 2 2 0 2.0000 -21694.8056
< 3 2 1 6.0000 -18018.6533
< 4 3 0 2.0000 -5477.1855
< 5 3 1 6.0000 -4457.4018
< 6 3 2 10.0000 -3574.4893
< 7 4 0 2.0000 -1394.8847
< 8 4 1 6.0000 -1061.8959
< 9 4 2 10.0000 -727.6412
< 10 4 3 14.0000 -373.0817
< 11 5 0 2.0000 -310.3843
< 12 5 1 6.0000 -207.1701
< 13 5 2 10.0000 -99.0042
< 14 6 0 2.0000 -48.0909
< 15 6 1 6.0000 -23.3716
< 16 5 3 3.0000 -3.4879
< 17 6 2 1.0000 -2.4702
< 18 7 0 2.0000 -4.3980
(Hartree a.u.)
total energy -27965.08138
kinetic energy 32916.01622
orbital energy -17704.54661
coulomb energy -70833.88056
hartree energy 10410.79449
exchange-correlation energy -458.01153
xc potential energy -608.27125
number of iterations 30 convergence 0.0E+00
integrated density 92.00000
... 1st derivative test 1 = 1.00000
... 2nd derivative test 1 = 1.00000
fhipp - all-electron atom done
=== TROULLIER-MARTINS mode === t
l n radius: node peak default core
x 0 7 1.739 3.444 3.000
x 1 7 0.000 0.000 4.000
x 2 6 1.068 2.388 3.000
x 3 5 0.433 1.042 3.000
=== pseudo atom ===
0 t 2.9748334 3.0483118 -4.3979630 -4.3979555
1.0000015 1.0000639
1 t 3.9867911 4.0852649 10.0000000 10.0000686
1.0000012 1.0000012
2 t 2.9748334 3.0483118 -2.4701861 -2.4701888
1.0000005 1.0000039
3 t 2.9748334 3.0483118 -3.4878567 -3.4879013
1.0000025 1.0000020
-- nlcv XC cutoff radius 0.99222
cv equidensity radius 2.76486
integrated model core density 33.12178
true core charge 86.00000
(Hartree a.u.)
total energy -45.43607
kinetic energy 12.61557
potential energy -22.67557
hartree energy 6.27051
xc energy -41.64657
integrated valence density 6.00000
... 1st derivative test 1 = 1.00000
... 2nd derivative test 1 = 1.00000
y range xvgr -11 1 3
fhipp - done for input @
@ 92.00 15 3 80.99E+00 : z nc nv iexc rnlc
@ 1 0 2.00 : n l f
@ 2 0 2.00
@ 2 1 6.00
@ 3 0 2.00
@ 3 1 6.00
@ 3 2 10.00
@ 4 0 2.00
@ 4 1 6.00
@ 4 2 10.00
@ 4 3 14.00
@ 5 0 2.00
@ 5 1 6.00
@ 5 2 10.00
@ 6 0 2.00
@ 6 1 6.00
@ 5 3 3.00
@ 6 2 1.00
@ 7 0 2.00
@ 3 t : lmax s_pp_def
@ 0 0.00 0.00E+00 t : lt rct et s_pp_type
@ 1 0.00 0.37E+00 t
@ 2 0.00 0.00E+00 t
@ 3 0.00 0.37E-03 t
Improved Polynomial Pseudocore -- dnlcc7
Pseudocore matching at i 394 r 0.992223523834322
Model parameters:
Pseudocore matching at i 394 r 0.992223523834322
Model parameters:
a0 196.747415814377
a1 0.000000000000000E+000
a2 0.000000000000000E+000
a3 -1359.98984837531
a4 2077.80605353870
a5 -1056.36746480728
a6 160.737503969336
Checking values at pseudocore radius:
19.6180132502683 19.6180132502683 density
-89.8716425001900 -89.8716425001908 1st derivative
486.490583948979 486.490583948972 2nd derivative
33.1217812898374 total pseudo charge
0.999999162920190 1st derivative test, should be = 1
1.00000116499618
2nd derivative test, should be = 1 (test may be inaccurate)
