Author Topic: COSMO frequency calculations in tmoleX  (Read 10093 times)

wwwups

  • Jr. Member
  • **
  • Posts: 10
  • Karma: +0/-0
COSMO frequency calculations in tmoleX
« on: February 25, 2008, 08:16:01 pm »
hi,
I am trying to do a UB3LYP/aug-cc-pvTZ frequency calculation using COSMO solvation for CO2- radical anion in water using tmoleX. I get this error in my project directory in the file aoforce.out "

" 2nd derviative of COSMO part and COSMO CPHF not implemented
 cosmo is accepted with numerical frequencies only
 ("frequency analysis only" in $drvopt group)

 MODTRACE: no modules on stack

 COSMO not implemented
 force ended abnormally
 2nd derviative of COSMO part and COSMO CPHF not implemented
 cosmo is accepted with numerical frequencies only
 ("frequency analysis only" in $drvopt group)

 MODTRACE: no modules on stack

 COSMO not implemented
 force ended abnormally"

How can I get tmoleX to do numerical frequencies ?
Thanks,
Pradeep

uwe

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 447
  • Karma: +0/-0
Re: COSMO frequency calculations in tmoleX
« Reply #1 on: February 26, 2008, 04:30:32 pm »
Hi,

Quote
How can I get tmoleX to do numerical frequencies ?

the current version does not support numerical frequency calculations, but you can run a usual single-point calculation, switch to the command line, change-dir to the directory where your job did run and start NumForce by hand (NumForce -help for details about the options you need). Make sure that you have set $TURBODIR to the TURBOMOLE installation path and expanded your PATH to $TURBODIR/scripts and $TURBODIR/bin/`sysname` as given in the documentation.

NumForce is a shell-script. Under Windows you will have to run it under cygwin.

Regards,

Uwe

wwwups

  • Jr. Member
  • **
  • Posts: 10
  • Karma: +0/-0
Re: COSMO frequency calculations in tmoleX
« Reply #2 on: February 28, 2008, 06:22:56 am »
Uwe,
Thanks for the reply. I can run NumForce calculations now. (NumForce works without the ri flag in control file.) When I try NumForce -cosmo, the program stops with
"presently -cosmo is implemented for ridft groundstate only". This is funny, because I got the same error message for a CO2 in air system as that of CO2- in air. The ground state of CO2 is a neutral singlet. Is there a reason why I am getting this error message for CO2 in air ?
Thanks,
Pradeep
p.s. When I mentioned that NumForce works without the ri flag, how comparable are the energies from RI-B3LYP calculations to normal B3LYP calculations? (I am using RI-DFT to optimize CO2 in COSMO model for vacuum . I now want to obtain the zero point energy but NumForce cannot do ZPE with RI.)

uwe

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 447
  • Karma: +0/-0
Re: COSMO frequency calculations in tmoleX
« Reply #3 on: February 28, 2008, 10:45:38 am »
Hi,

the options of NumForce do not switch on RI or COSMO, you have to add -ri if RI is used in the input file.

The question is: what do you want to calculate?? Second derivatives and frequency calculations are two different goals and need different steps:

  • second derivatives to check if your geometry is a minimum or a transition state

        In that case just run NumForce without the -cosmo option. This will work with and without RI.


  • frequency calculations to get a spectrum to compare with experiment

        In that case the theory is a little bit more difficult since not all contributions that come from the solvent should be included in the frequency calculation. Here you have to give the -cosmo option, and the charges on the COSMO cavity will be frozen, and some other approximations will be used for the calculation. This feature is only implemented in the binaries that are capable of doing RI - and that is the reason why NumForce with -cosmo requires an RI run.

Hope this helps,

Uwe

wwwups

  • Jr. Member
  • **
  • Posts: 10
  • Karma: +0/-0
Re: COSMO frequency calculations in tmoleX
« Reply #4 on: February 29, 2008, 07:31:57 am »
Hi,

the options of NumForce do not switch on RI or COSMO, you have to add -ri if RI is used in the input file.

The question is: what do you want to calculate?? Second derivatives and frequency calculations are two different goals and need different steps:


  • frequency calculations to get a spectrum to compare with experiment
Thanks, NumForce -ri -cosmo worked!
[/list]