I'm calculating the energy of all the E isomers of octadecene using Opt+Freq with DFT B3LYP 3-21G. Everything was fine from 1-octadecene to E-5-octadecene, but from E-6-octadecene onwards it wasn't able to optimize the molecule and it started turning out imaginary frequencies. I don't know what changed between 5 and 6, could it be that the hydrocarbon chain is too large to optimize? Does anyone know what could be causing thus and how to correct it? Help is very much appreciated.
Minimum energy structures should have no imaginary frequencies. With your structures I expect some I. f are due to methyl rotation. Check the output and if that is so, try to rotate the molecule again and reoptimze
Also, your basis set is too small. You should make it bigger
Just an addition to Ismail Badran answer, sometimes geometry optimisation doesn't end in a minimum but in saddle point of 1st order, where the gradient vanishes too. The more complex the structure, the more likely this may happen.
To get a clearer information, which group might be affected, have a look at the normal modes/vibrations and then follow the normal mode.
If you don't have to stick with gaussion, try using the orca quantum chemical software, which has this outlined approach implemented.
And as Ismail Badran mentioned, try use some larger basis set and include dispersion correction. Van-der-Waals forces should not be neglected even for small systems.
Hübler answer is pretty complete already, but I will clarify only one thing: 1st order saddle points, most often associated to a transition state structures between reactants and products, might also appear between two local minimum conformers. Cyclooctatetraene (COT), because not aromatic (4n pi electrons instead of 4n+2), is not expected to be planar. So COT planar might be a stationary point, but a saddle point between two non planar conformes. This might be the case with your molecules. By viewing the normal coordinate, as suggested, you will be able to check this hypothesis. Regards,
Please try fixing some of the dihedral angles (one must make sure that one is not fixing any such variabls that must have ther freedom to change, but there are some other trivial movements that could be prevented) etc., and re-optimizing.
Unnecessary flip-flops of a long or flexible molecule can keep giving imaginary frequencies even if you try to follow the largest imaginary frequency and re-optimize.
- Badges
- Science topic
- Similar topics
- Academic Journals