olgerdovich
olgerdovich@yandex.ru
>The calculations were perfomed at b3lyp/LANL2Zdp level.
>The structures are only different on the n-propyl substituent orientation ( bent and stretched).
>
>I have discussed that energy difference between the geometries caused by hydrogen bonds and steric effects.
>I hope the attached file might help to understand..
>best regards,
>Lucas
It appears that you should use D3-correction (http://onlinelibrary.wiley.com/doi/10.1002/jcc.21759/abstract) which significantly improves discription of non-covalent interactions. As figure 4 of the cited paper shows, B3LYP accuracy for non-covalent interactions is improved from 4 kcal/mol to approx. 1 kcal/mol. Note that this correction is "cheap" and goes with almost no computational cost. However, your difference in energies (if it remains after D3-correction) is very close to this error.