I have found in literature many different combinations of functionals and basis sets with the respective correction factors. Which is the best combination among all of them? I am interested in vibrational frequencies of CO, CN, OH and other small ligands attached to Iron, Nickel (for example CN stretching connected to Fe or Ni around 2000-2100 cm-1).
B3LYP with 631G** works well for me,However I have calculated vibrational frequencies for organic molecules only.
IR and Raman shifts can be satisfactorily reproduced by use of B3LYP. However for best results you should use triple zeta basis set (e.g. 6-311G**). If you have large system and you are forced to use smaller basis set (e.g. 6-31G**), you can multiply the wavenumber value by the scaling factor that can be found in the literature.
[A] In Fe(CO)5, Fe2(CO)9,Fe3(CO)12, Ni(CO)4 and Ni2(CO)8; Fe(II) or Ni (II) with CN--- DZ worked both in IR&Raman.
Pls. note that all these complexes were diamagnetic .
[B] In IR( NOT RAMAN) of Ni(II) & CN again DZ worked.
[C] In IR( NOT RAMAN) of Fe (II) & CN – GGA;BP/ SZ worked.
i agreed with Patryk Wlodarczyk. i tried organic molecules with different methods and basis set combinations. But DFT/B3LYP with 6-311++G(d,p) only yields better results when compared to that of the experimental.
inorganic materials (like NaMgH3) I have used ABINIT code (plane wave+ pseudpotentiel)) see our paper in physica B Volume 406, Issue 4, 15 February 2011, Pages 1000-1003
Transition Metal cofactors: BP86//TZVP/6-31g*
No need of scaling factors.
- Badges
- Science topic
- Similar topics
- Chemistry, Organic
- Organic