The problem here is that you're most probably using a single, isolated molucule of water for calculations. No wonder that you correctly get only 3 vibrations. You can think of it as simulating water in gas-phase (vapor) and the experimental spectrum of such system would be kind of similar.

The experimental spectrum of liquid water or ice is completely different due to contributions from hydrogen bonds between water molecules. I have no experience in modelling vibrot spectrum of water but I would imagine that in order to get something similar you would have to create a model water cluster (there are many computational papers describing such clusters of different sizes) - the larger, the better. 

Also: please read this: http://www1.lsbu.ac.uk/water/water_vibrational_spectrum.html

Mr. Stoicu,

As Mr. Trzaskowski has correctly pointed out within the THz-region you can obain mainly collective vibrations of ensemble of interacting species (ions, molecules). Please find as attachment a computation of 5H2O cluster with few visualized vibrations.  

Thank you for your answer! I'm already running the computation with 5 water molecules, as you suggested, for a week now, so I'm still waiting...

I have another question: 

When computing frequencies, a scaling factor is required in order to be able to compare the experimental with the computed spectra. Given the fact that freq=vibrot is also a frequency computation I'm guessing that a scaling factor/function is also necessary. Correct me if I'm wrong!

Does anyone have a reference for a paper where scaling factors/functions appropriate for CCSD with 6-311G(d,p) are duscussed!?!?!? - the most recent paper I have about scaling factors/functions discusses DFT with B3LYP and several basis set configurations.

Hello Alexandru,

I am trying to get the THz spectra of a single polyurea chain but I don't know how to do it using Gaussian. Do you have a step-by-step guide on how to extract that information from the simulation?

Thank you,