eiger function not working

[golden]

Hello,
  after optimizing a S6 symmetry molecule , I changed the the charge from -1 (closed shell) to -2 (open shell) changing the control files, occupation numbers where

as in -1 MO from eiger would say

Code: [Select]

Number of MOs=    782, Electrons=    782.00, Symmetry: s6   
 
   Nr.   Orbital    Occupation       Energy
  782. b 130 eu       1.000        -0.160135 H =        -4.357 eV
  781. b 129 eu       1.000        -0.160135 H =        -4.357 eV
  780. a 129 eu       1.000        -0.160135 H =        -4.357 eV
  779. a 130 eu       1.000        -0.160135 H =        -4.357 eV
  778. b  65 au       1.000        -0.162460 H =        -4.421 eV
  777. a  65 au       1.000        -0.162460 H =        -4.421 eV

where corresponding control file would have;

Code: [Select]

$alpha shells
 ag      1-66                                   ( 1 )
 eg      1-130                                  ( 1 )
 au      1-65                                   ( 1 )
 eu      1-130                                  ( 1 )
$beta shells
 ag      1-66                                   ( 1 )
 eg      1-130                                  ( 1 )
 au      1-65                                   ( 1 )
 eu      1-130                                  ( 1 )

So when I changed the charge to -2  the control file would be like;

Code: [Select]

$alpha shells
 ag      1-66                                   ( 1 )
 eg      1-130                                  ( 1 )
 eg  131-132  (1/2)
 au      1-65                                   ( 1 )
 eu      1-130                                  ( 1 )
$beta shells
 ag      1-66                                   ( 1 )
 eg      1-130                                  ( 1 )
 au      1-65                                   ( 1 )
 eu      1-130                                  ( 1 )

after changing the charge when I try to run eiger it would give the error message as ;
<rdlist> : index =   131 lies not within (    0,  130)

 
 MODTRACE: no modules on stack

  fatal <deforb> error
  ended abnormally
  ended abnormally
program stopped.


(if I run the ridft it would not run through)
is there any thing that I could do to run the calculation with out decreasing the symmetry of the molecule?

I would appreciate if you could shine some light on my problem.

Thank you very much in advance.

[antti_karttunen]

Hi,

Just to clarify the situation: Did you generate the initial EHT guess for the open shell (charge -2) system with Define? Or did you just manipulate the occupation numbers in the control file after generating the guess for charge -1? The EHT guess only yields the occupied orbitals, so you cannot manually add electrons into the virtual space before you run ridft/dscf.

Best,
Antti

[golden]

Hello,
  thanks for getting back to me.

yes I intially created the -1 charge(closed shell ) system with define , as a openshell with alpha and beta electrons , and then changed the occupation numbers in the control file so I could change it to -2.   So is this wrong?

Because I used to do this to change the charge from -1 --> 0--->+1 but then again I was always removing an electron. So I thought I could do the same to go from -1--> -2 .

I found it confusing to create -2 charge with using define program. because it always keeps giving me an triplet state, but when I was able to change it,I couldn't figure out to which state it changed to. !!!  Is there any way to conform to which state did i changed it to? (singlet, triplet... etc..)

thanks.

[antti_karttunen]

Hi,

Yes, you cannot add electrons to the system if you only have the EHT guess orbitals. You could run dscf/ridft for the singlet state (using UHF) and then add electrons. But using define should work, as well. If you know, which spin state you are aiming for, you should be able to set this already in the "OCCUPATION NUMBER ASSIGNMENT MENU". Just play with the choices u and uf:

u <int>   : CHOOSE UHF WITH <int> UNPAIRED ELECTRONS
uf <int>  : AS u, BUT ALLOW FRACTIONAL OCCUPATION NUMBERS

Since you were using fractional occupations in your original message, "uf" might be what you are looking for.

Hope this helps,
Antti

[golden]

thanks for the heads-up,  I will try to play around that. Just curious is there is an better way to give the guess other than using eht?

thanks again for the help.

[antti_karttunen]

Not really. You could always read in the MOs from a lower level calculation, but you would still need to use EHT to get an initial guess for that one