How to correlate Theoretical IR peak half with and half height with Experimental

Sakib

In the experimental IR, there is a broad peak in some cases. Where in theoretical Ir the peak is so sharp. when we extract data or visualize via gauss view we can adjust the IR peak half with and half height value in gauss view where the default value is 4. Now my question is how can I conclude which value to choose in the gauss view so that my theoretical Ir value looks close to the experimental value? value and the epsilon value also correlate?
I have attached one of my experimental and calculated IR spectra.

wzkchem5

The IR peak widths are primarily determined by (1) the natural width (determined by the lifetime of the vibrational excited state), (2) the presence of many different conformers/aggregates whose IR signals are slightly different and sum to a wider peak than each of the individual peaks, and (3) solvation effects. (3) is a special case of (2) since a solution is nothing but an aggregate of the solute and a lot of solvent molecules.
The broadening functionality of GaussView primarily models (1), that's why the peaks are very narrow by default and do not reflect the experimental differences of peak widths. To quantitatively reproduce the experimental widths, you have to account for (2) and (3), which can be done by either (a) calculating the IR spectra of all important conformers of your molecule and Boltzmann-average them, or (b) do an MD simulation and extract the IR spectrum from the autocorrelation function (https://brehm-research.de/spectroscopy.php).
Nevertheless, for studies that only involve IR peak positions but not the peak widths, it is usually acceptable to not reproduce the experimental peak widths. If your study does not rely on the peak widths at all, then you can simply move on.

Sakib

Thanks for your reply. I understand your explanation but if still want to broaden the peak what should I need to take into consideration? given that the computed spectra are extracted from gas phase calculation.

Another thing is I have already simulated this complex. Can I analyze the trajectory with Travis? to get the IR spectra via  the trajectory xtc file?

wzkchem5

Sakib 发表于 2022-9-15 15:51
Thanks for your reply. I understand your explanation but if still want to broaden the peak what shou ...

As I said, either you calculate the IR spectra of all important conformers and take the Boltzmann average, or you use programs like TRAVIS to extract the IR spectrum from an MD trajectory instead. Either method will give you a spectrum where all peaks are broadened to a proper degree.
I don't know enough about TRAVIS to answer your second calculation. But I guess that even if you cannot directly analyze your trajectory with TRAVIS, there may be some way to convert the trajectory to the correct format so that TRAVIS can recognize it.

Sakib

Thanks for your time, dr Zikuan Wang.