I have carried out geometry optimizations and frequency calculations using Gaussian for a number of weak complexes. It has been found that different theories give significantly different minima. But only one particular level (suppose XX) has given a minimum, the frequency of which is quite comparable with the experimental frequency. The frequency shifts of the other levels are comparable to each other, but significantly differ from the experiments.
So, my query is, can I fix the particular level of theory (XX) for comparing my experimental findings, or should I believe the other levels?
Hello
I would have stopped there (theory (XX)), especially if the calculation was too costly
Dear Shubhra Sarkar
It happens in many systems such as quantum dots semiconductors. For example PBE functional underestimates the band gap while B3LYP shows better conformity with experiments. However, B3LYP is costly and it would take too much time but PBE computational cost is much more reasonable for molecular dynamics. Even in this case, I have seen papers comparing their results with experiments while confessing this fact. So, it would highly depend on your case of study. If you have obtained a good conformity with experiments then you have found the right choice.
- Badges
- Science topic