PC GAMESS/Firefly-related discussion club

Learn how to ask questions correctly

Alex Granovsky
gran@classic.chem.msu.su

Could you please provide more details on how exactly
the BDE values were calculated? E.g., did you allow
the geometries of two radicals to relax? Did you
performed three separate calculations (one for initial
molecule, second for radical and third for the rest of
molecule)?

Note, almost all (approximate) XC functionals have
some systematic errors when going from singlet state
to states with higher multiplicity. To check your DFT
numbers, you can try to perform MP2 calculations at
geometries obtained at DFT level.

Regards,
Alex Granovsky


On Mon Dec 8 '08 10:44am, James Phil wrote
------------------------------------------
>I would like to ask about the consistency of using DFT methods in PCGAMESS to calculate Bond Dissociation Energy of Different Molecules. IM currently using B3LYP/6-31+G(3d,p,f) as my chosen method and basis set and I found out it computes the BDE Values of some compounds reasonably only if the involve radical is Hydrogen (H). However, when I tried to compute the BDE Values of compounds whose dissociation reactions involves other radicals aside from hydrogen (e.g CH3, SiH3), there is significant error in the results. There's a difference of about 6-10 kcals/mole with the computed values with the literature values. What could be the reason for this? Im looking at the capability of DFT methods on integrating electron correlation between molecules. Im also considering the size of the radicals formed. Am I at the right path? Please do help.