I have finished the NEB calculation, but it seems all the energy I got is Internal energy instead of Gibbs free energy, so do I need to do Gibbs free energy modification? Is there any standard procedure for it?
1
-
1$\begingroup$ I'm not familiar with VASP but in Gaussian there is a standard way to calculate this for a single molecule by performing a frequency calculation. See google.com/url?sa=t&source=web&rct=j&url=https://… $\endgroup$– Hayden SApr 23, 2022 at 20:08
Add a comment
|
3 Answers
$\begingroup$
$\endgroup$
To get the Gibbs Free energy from the internal energy, you have to consider the configurational entropy. I anticipate you need the Gibbs Free energy to calculate the activation barrier as a function of the temperature. There are two ways to calculate the activation barrier as a function of temperature, namely, molecular dynamics, and NEB followed by kinetic Monte Carlo. Details of these methods can be found in section 4 of this review.
$\begingroup$
$\endgroup$
I think the first question you have to ask is whether you can get an accurate free energy for your reaction barrier in VASP. If you have a reaction with a small molecule on an oxide surface under vacuum, the configurational entropy may be insignificant. In this case, a vibrational analysis (correcting for ZPE and possibly vibrational entropy) at the transition state may be appropriate. The total correction will likely be small in this case.
For a vibrational calculation, you can just use IBRION=-5 and NFREE=2 as shown on the VASPwiki. For the TS, you should get a single nonreal frequency because you're at a saddlepoint. You can use these frequencies to get the ZPE correction and the vibrational entropy from the harmonic oscillator equations. I'll reproduce the harmonic oscillator entropy equation here: \begin{equation} S = -Nk_{b}(1-e^{-\beta \epsilon }) + Nk_{b}\frac{\beta \epsilon }{e^{\beta \epsilon } -1} \end{equation}
On the other hand, for a molecule in solution or a solvated interface, simple vibrational analysis will not describe the entropy of the system very well. Unless you're willing to use a more advanced sampling method, you may just want to report the enthalpic barrier.
$\begingroup$
$\endgroup$
It depends on why you've made the NEB. If to find a possible reaction coordinate, you can use the ∆E as an indicator.
Another thing is that you can't perform the "traditional" thermochemistry correction (so diagonalizing the Hessian matrix) because you aren't on a stationary point. I know there are some methods, but I'm not an expert on this so unfortunately, I can't help you by pointing out some literature.
answered Apr 26, 2022 at 17:59