Author Topic: Generate WFN file  (Read 16955 times)

scope

  • Jr. Member
  • **
  • Posts: 18
  • Karma: +0/-0
Generate WFN file
« on: March 01, 2007, 01:40:53 PM »
Hi,
I just stumbled across a very old post by Uwe post in an abandoned mailing list  ;D
Just thought I´d share this information because it is not in the manual (and I had a hard time before I figured this out...).
If you need Gaussian98-style WFN files (e.g. for Bader Analysis with Xaim or Proaim), you just have to put the keyword $wfn in your control file (job must be finished), then run ridft-proper, rimp2-proper or whateveryouused-proper.
It generates a file named "wfn", which you can then plug into your analysis program.

Have fun,
Nils

christof.haettig

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 291
  • Karma: +0/-0
    • Hattig's Group at the RUB
Re: Generate WFN file
« Reply #1 on: March 28, 2007, 10:08:57 PM »
Yes, that's true. But it is based on a old version of Baders AIM programs, where
the order of the cartesian components of the basis functions was only defined
up to f functions. For higher angular momenta the order on Turbomoles wfn
file might be different from what Gaussian uses!
Christof

christof.haettig

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 291
  • Karma: +0/-0
    • Hattig's Group at the RUB
Re: Generate WFN file
« Reply #2 on: June 02, 2008, 04:44:03 PM »
I tested the WFN routines with a copy of the Bader AIM programs which used (up to d functions) the canonical order, which is also the order Turbomole uses internally for the cartesian basis functions. So, the routine which generates the WFN file in Turbomole uses the canonical order of the cartesian for all quantum numbers. Currently there is no option to write the WFN file in spherical basis function format. I don't have a description of a WFN format for sphericals.

Regards,
                       Christof