Author Topic: CCSD(T) calculation on top of SCS-MP2 optimized calculation  (Read 882 times)

mamun006

  • Newbie
  • *
  • Posts: 4
  • Karma: +0/-0
CCSD(T) calculation on top of SCS-MP2 optimized calculation
« on: October 18, 2016, 04:36:26 pm »
I ran some SCS-MP2 geometry optimization on some carbohydrates geometry. Now, I want to run CCSD(T) single point calculation. So, I ran the define again and in the last menu I switched on ccsd(t) model. Then I ran a dscf calculation followed by a ricc2 calculation. Is it a right approach? My energy file is showing HF energy and MP2 energy. Also, my gradient file is showing HF+MP2 energy. I am confused. It's not showing any ccsd(t) energy. However, in my control file, it's still showing $ricc2 mp2 energy only. Should I change it manually to ccsd(t)?  Also it would be really helpful if someone can provide a step by step procedure to run a ccsd(t) single point calculation. This section (cc2) in the manual is very confusing. Thanks for the help.

gunnar.schmitz

  • Developers
  • Jr. Member
  • *
  • Posts: 10
  • Karma: +0/-0
Re: CCSD(T) calculation on top of SCS-MP2 optimized calculation
« Reply #1 on: October 19, 2016, 09:38:36 am »
Dear mamun006,

in principle you did the correct setup. If you rerun define you have to do in the last menu
  • cc
  • ricc2
  • model ccsd(t)

If in you control file

$ricc2
  ccsd(t)

does not show up, something went wrong. Maybe it is a save choice to contact the Turbomole support since you probably found a bug in define. Nevertheless you can change the datagroup by hand and do your CCSD(T) single point calculation. If there is still something like geoopt model=mp2 you should also delete this line.
Afterwards it depends which Turbomole version you are using. Since version 7.0 the functionality for doing CCSD(T) calculations was moved from the ricc2 program to the new ccsdf12 program. So you have to use this instead if you are using TM-V7+

Best wishes,
Gunnar

mamun006

  • Newbie
  • *
  • Posts: 4
  • Karma: +0/-0
Re: CCSD(T) calculation on top of SCS-MP2 optimized calculation
« Reply #2 on: October 20, 2016, 08:13:27 pm »
Dear Gunner,

Thanks for the answer. Somehow, define is unable to change mp2 to ccsd(t). Now, I changed it by hand and also I had to change MPI to SMP. Thanks!