Hello everyone!
I'm currently computing a quite big potential energy surface for a catalytic transformation. I'm using BP86/6-31G**/LANL2DZ for gas phase optimizations and M06L/6-311G**/SDDALL for single point calculation in solvent with and without GD3 dispersion correction. In all the cases (intermediates and TSs) I am comparing the zero-point corrected energies, enthalpies and Gibbs free energies. The issue I'm currently facing in all the methods I used is that in two small parts of the profile there is a huge difference between ZPCE and free energies but not between ZPCE and enthalpies. In these two cases I have comparison of energy like A+B compared to C but this is the case also in other parts of the mechanism where I did not encountered the same problem. I even tried to calculate the basis sets superimposition error but that was found to not be the problem.
Does anyone has suggestions on how to approach this situation and where/how should I find the reason for such difference?
Thanks a lot!
SInce the only difference between enthalpy and Gibbs free energy is the entropy, this is the likely culprit. In A+B = C type reactions the contribution of the entropy change is likely to be substantial (particularly in higher temperatures), though obviously depending on the structure of reactants and products. I would say that this is not really surprising for reactions where there is a major change in bonding pattern and conformations of the products versus reactants, though obviously the experimental validation would be very welcome for such or similar case.
That doesn't sound strange to me. ZPE correction is only dependent on the PES curvature difference between reactant state and transition state (i.e. difference in second derivative matrices at the two stationary points). Low ZPE contribution means that the 3N-7 vibrational frequenceis of TS is comparable to the 3N-6 frequences of RS. There are many unrelated sources of entropy differences.
Mr. Franzoni,
You could evaluate the differences of the values using 'freq' test.
Hallo, Ivan!
Zero-point corrected DFT energy has the meaning of the Gibbs free energy at absolute zero temperature (= enthalpy at absolute zero).
Gibbs free energy (and enthalpy) has physical meaning only at local minima of potential energy surfaces. It is sometimes used also for transition states but the correction on imaginary vibrations is necessary in such cases.
Best wishes
mb
Mr. Franzoni,
In context the answer provided by Mr. Breza you couls add to the route section temperature=273.15. Thus teh freq-data are obtained at T=0oc.
DFT is just an approximation of the Born-Oppenheimer ground state approximation of the interacting electron gas in the field of the nuclei at T=0.In a dilute molecular gas,it may yield reasonable enthalpies and if the vibrational spectra of the ground states involved are calculated first order corrections to the free energies.But the complete QM treatment of a many body molecular problem that would yield reliable free energies (Helmholtz,Gibbs etc) is music for the future and should include excited states as well via some density matrix generalizations.Horrible stuff.
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry