Hi,
In short I am getting a bit overwhelmed with the difference in nomenclature used by the Gaussian output, text books, and people's suggestions.
I am trying to pin point the most energetically favourable conformation between two geometry optimised structures (A and B). Both structures are in a minumum with no imaginary frequencies. What 'value' would be the best to report. Here is an example of where I am looking in a Guassian log file after a frequency calculation.
Structure A)
Zero-point correction= 0.549237 (Hartree/Particle)
Thermal correction to Energy= 0.580114
Thermal correction to Enthalpy= 0.581058
Thermal correction to Gibbs Free Energy= 0.486170
Sum of electronic and zero-point Energies= -1484.656742
Sum of electronic and thermal Energies= -1484.625865
Sum of electronic and thermal Enthalpies= -1484.624921
Sum of electronic and thermal Free Energies= -1484.719809
Structure B)
Zero-point correction= 0.549667 (Hartree/Particle)
Thermal correction to Energy= 0.580703
Thermal correction to Enthalpy= 0.581647
Thermal correction to Gibbs Free Energy= 0.483468
Sum of electronic and zero-point Energies= -1484.647529
Sum of electronic and thermal Energies= -1484.616493
Sum of electronic and thermal Enthalpies= -1484.615549
Sum of electronic and thermal Free Energies= -1484.713728
I should then (hopefully) be able to report on the difference of X (my question) between structure A and structure B.
Thanks for your help. I'm still very new to ab initio and I am far more used to Free Energy MD calculations.
The 1st structure has lower Sum of electronic and zero-point Energy and lower Sum of electronic and thermal Enthalpies. Thus, the 1st stucture is the lowest energy isomeric form. The energy difference betweeen them can be also substantial when computing thermochemistry, as well as you have, in fact, low-lying isomeric form of the 1st structure
The 1st structure has lower Sum of electronic and zero-point Energy and lower Sum of electronic and thermal Enthalpies. Thus, the 1st stucture is the lowest energy isomeric form. The energy difference betweeen them can be also substantial when computing thermochemistry, as well as you have, in fact, low-lying isomeric form of the 1st structure
It depends how big molecule you talk about, and what experimental results you would like to compare to. Your ultimate goal would be the relative free energy, that is Sum of electronic and thermal Free Energies. But the entropic part is not that reliable as it is most often uses a harmonic approximation, and lacks terms arising from conformational sampling. It is ok to use DeltaG though, just you should know what terms your values consist of.
I agree with Anthony Nash though, that safest is to consider DE and DH, especially that you did not mention what molecule you calculate for.
Most often you report in relative energies either using kcal/mol or kJ/mol, that is (A-B)*627.51 for kcal/mol, or (A-B)*627,51*4.186 for kJ/mol
In case you want to have a very good /publication quality value, you almost always have to calculate electronic energies on higher level of theory. Then, you could add the ZPE and Enthalpic, etc corrections to that energy and not perform freq calcs on the higher level. See the white papers on thermochemistry for G09 and the section Output from a frequency calculation on which term means what.
Hope this helps
Hi Tamás Beke-Somfai, thanks for that information. Especially regarding what energies to measure if recalculating on a higher level of theory.
You often can't go wrong with just reporting the total energies. Though I am sure you already know, make sure that you are comparing energies computed at the exact same theoretical level.
The answers that have been given already might be a little bit more accurate, but the total energy is probably the most reliable value in the output anyway...
by looking to sum of electronic and zero point energies, so which one has the largest negative value is the most stable (structure A)
- Badges
- Science topic