[molpro-user] UHF convergence failure, 2010.1 version

[Neeraj Rai]

Dear Molpro users,

  I am trying to run UHF scf calculation on NpO2^+ but it doesn't seem to
get beyond 2-3 iterations and the program quits with the following message:


 Orbital guess generated from atomic densities. Full valence occupancy:
14   7   7   3

 Molecular orbital dump at record        2200.2

 Initial alpha occupation   6   5   7   3
 Initial beta  occupation   6   5   5   3

 ITERATION    DDIFF          GRAD             ENERGY
2-EL.EN.            DIPOLE MOMENTS         DIIS
    1      0.000D+00      0.000D+00      -561.57652943   1166.824804
-16.24030    0.00000    0.00000    0
 Occupation changed:
 Alpha spin:  11   4   4   2
 Beta  spin:  11   4   4   0
    2      0.302D+00      0.100D+00      -509.23245161    780.797020
55.57348    0.00000    0.00000    1
 Occupation changed:
 Alpha spin:   5   5   6   5
 Beta  spin:   5   5   6   3
    3      0.309D+00      0.185D+00      -518.38899029   1312.399657
-23.94045    0.00000    0.00000    1
 Occupation changed:
 Alpha spin:  11   5   4   1
 Beta  spin:  11   3   4   1
    4      0.671D+00      0.202D+00      -344.88378829    719.421186
92.86459    0.00000    0.00000    1
 Energy increased. Restoring previous occupations:
 Alpha spin:   5   5   6   5
 Beta  spin:   5   5   6   3

 IOCC
   1-  4     5     5     6     5

 ICLOS
   1-  4************************
 ? Error
 ? Inconsistent closed shell occupation
 ? The problem occurs in uhf_occ

 GLOBAL ERROR fehler on processor
0




Any suggestions how I can circumvent this problem? The input is as follows:


***,CH4-CH4
memory,200,m
GTHRESH,ENERGY=1.d-8
geomtyp=xyz
geometry={                 !dummy center in center of mass
3
Dimer
Np     0.000000    0.000000   0.000000
O    -1.739061    0.000000   0.000000
O     1.739061    0.000000   0.000000
}
set,charge=+1
set,spin=2
basis={
spdf,O,avtz;C;
ecp,Np,ecp60MWB;
spdfg,Np,ECP60MWB_ANO;C;
}
uhf;



Thanks.


-- 
Regards,
Neeraj.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20110216/c5eeda80/attachment.html>