Author Topic: Dummy atom  (Read 12228 times)

olehtone

  • Newbie
  • *
  • Posts: 5
  • Karma: +0/-0
Dummy atom
« on: February 22, 2007, 09:44:37 am »
Hi,

I would like to do a potential energy surface scan for a linear molecule and I have a dummy atom in the Z-matrix. Tmole script doesn't seem to understand x (or q) as a dummy atom. Is there any solution for this?

Olli

uwe

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 425
  • Karma: +0/-0
Re: Dummy atom
« Reply #1 on: February 22, 2007, 11:28:48 am »
Hi Olli,

tmole in Turbomole version up to 5.8 had problems with dummy atoms in the Z-matrix. That has been fixed in Turbomole 5.9 (at least for 'q', but not for 'x' I guess).

Regards,

Uwe

lsikk

  • Newbie
  • *
  • Posts: 9
  • Karma: +0/-0
Re: Dummy atom
« Reply #2 on: March 17, 2007, 12:21:07 pm »
Hi

Do the versions up to 5.9 have problem with dummy atoms in general or is it just Tmole problem?
I prepared my input with define and RIDFT stops, telling me that q atoms dont have a basis set
$atoms section in control file looks like this:
$atoms
c  1-3,8                                                                       \
   basis =c TZVP                                                               \
   jbas  =c TZVP
h  4-6,10                                                                      \
   basis =h TZVP                                                               \
   jbas  =h TZVP
q  7,9                                                                         \
   basis =none
« Last Edit: March 17, 2007, 12:22:47 pm by lsikk »

christof.haettig

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 258
  • Karma: +0/-0
    • Hattig's Group at the RUB
Re: Dummy atom
« Reply #3 on: March 18, 2007, 11:45:22 pm »
This has is a  general problem in DFT calculations in Turbomole:
The DFT  code selects the grid for the numerical integration based
on the atom labels in the basis sets. Thus with a 'none' basis the program
aborts because is can not determine which grid it should use...
(Maybe the DFT people can comment on the reason for this choice!?)

Christof

uwe

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 425
  • Karma: +0/-0
Re: Dummy atom
« Reply #4 on: March 19, 2007, 04:18:22 pm »
Hi,

in your case, ridft complains about missing basis sets because your dummy atoms do not have a jbas= line in the $atoms section. Since ridft tries to read in the auxiliary basis set for all atoms, including dummy atoms, it will stop at that point.

Just try to change your $atoms section by adding a backslash to the end of the last line of the q section, and add the line:

jbas =none

Sould look like:

q  7,9                                                                         \
   basis =none                                     \
   jbas = none

Regards,

Uwe

P.S.: I have mentioned above that tmole in Turbomole 5.9 can deal with dummy atoms in z-matrix inputs - but what it does is simply not to write the dummies to the coord file... usually those dummies are just needed to get a proper z-matrix, so in most cases that fix will work. If you want to keep the dummies, you will have to do the input with define and/or by hand, i.e. without tmole.
« Last Edit: March 19, 2007, 04:22:48 pm by uwe »

lsikk

  • Newbie
  • *
  • Posts: 9
  • Karma: +0/-0
Re: Dummy atom
« Reply #5 on: March 21, 2007, 03:47:03 pm »
Thanks Uwe, adding jbas =none works fine.