CCL: Benchmarking the G09 calculations

[Mikael Johansson <mikael.johansson]![iki.fi>]

 Sent to CCL by: Mikael Johansson [mikael.johansson:_:iki.fi]
 Hello Ambrish,
You wouldn't necessarily ask a great computational chemist to perform
 a
 ten-step organic synthesis; [s]he might be good at wet chemistry as
 well,
 but you don't know that.
So instead of benchmarking your method against a property you're not
 actually interested in, you should benchmark for the right property. So
 choose molecules similar to your own that have available data for IP,
 EA,
 etc., and see how the method performs for those. Probably there is
 plenty
 of literature on what methods are appropriate for your systems already
 as
 well.
Nothing wrong with proving the reliability of your method for other
 properties as well, of course.
 Cheers,
     Mikael J.
     http://www.iki.fi/~mpjohans/
 On Wed, 18 Sep 2013, Ambrish K Srivastava ambrishphysics- -gmail.com wrote:
> I want to calculate some electronic parameters viz. IP
>  , EA etc of
>  inorganic molecular species for which corresponding experimental data
>  are unavailable. Of course, I have experimental FTIR and Raman spectra
>  of my species. To benchmark my calculations of electronic parameters,
>  would it be good idea to match calculated IR and Raman spectra with
>  corresponding experimental ones? Can the method giving better agreement
>  between calculated and experimental wavenumbers be treated as more
>  reliable for the calculation of electronic parameters?
>  Any suggestions will be highly appreciated..
>  thanking you all in advance!
>  Ambrish K. Srivastava
>  CSIR JRF
>  Department of Physics
>  University of Lucknow