Author Topic: No saving of new coordinates of excited state.  (Read 4458 times)

Addiw7

  • Jr. Member
  • **
  • Posts: 10
  • Karma: +0/-0
No saving of new coordinates of excited state.
« on: February 13, 2013, 10:54:42 AM »
Hi,

I wanted to do ricc2 geometry optimization  calculations of first excited state on a molecule which ground state geometry has been optimized. Fristly I did dscf calculations to get the virtual orbitals. Than I made a new control file for ricc2 geoopt cc2 (a 1) optimization. As a coord file I took one I got from dscf calcualtions. Now, I run ricc2 and they are finished like in 30-60 minutes (depends on basis set), ricc2 says it is done, but the coords are still the same. What is possibly wrong? I made control file using both define and tmolex, still the same.

Thank you in advance.

Hauke

  • Full Member
  • ***
  • Posts: 37
  • Karma: +0/-0
Re: No saving of new coordinates of excited state.
« Reply #1 on: February 14, 2013, 02:19:42 AM »

Even if you have specified "geoopt cc2 (a 1)", if you called ricc2 you just did a single point + gradient calculation and therefore the coordinates don't change.

To optimise the structure you have to run
jobex -level cc2
(in contrast to the dft case where you have to use jobex -ex, for cc2 the $ricc2 and $excitations keywords in the control file decide if the ground or excited state should be optimised)

I'm not an expert in this type of calculation but I have my knowledge from the  tutorial where in chapter 9.6 is described how to get a excited state structure at CC2 level. I hope that it is correct and works.