A common way is to take the main Raman peak. You should consider the cation that is likely to exhibit the higher vibration energy, for example in oxides:

B> P> Si > Ge and Al> As> Te and Sb etc.

A rough estimation is also a matter of experience. In the same way as you can estimate the refractive index from the chemical formula, it is also possible to check the approximate value of the phonon energy just by lookin at its composition.

The anions also must be taken into account. Oxygen and fluorine have similar contributions. For chalcogenides, sulphur has a higher contribution to phonon energy tha Se and Te. In complex multicomponent glasses you should focus on lighter anions, and on cations that exhibit the larger "field strength".

Finally, as far as phonon energy is used to estimate the non radiative desexcitations in rare earth cations, you should also consider the local environnment of the rare earth. Unfortunately, it remains largely unknown. But it is possible to have some indirect information about it.

Last point: some impurities (hydroxyls) have also a significant influence.

Probably the second order Raman spectrum would help you to get the desired information.

But in a complex glass is not very obvious identified the 2nd order modes. Do you know any references for that??

Thanks a lot!!

A common way is to take the main Raman peak. You should consider the cation that is likely to exhibit the higher vibration energy, for example in oxides:

B> P> Si > Ge and Al> As> Te and Sb etc.

A rough estimation is also a matter of experience. In the same way as you can estimate the refractive index from the chemical formula, it is also possible to check the approximate value of the phonon energy just by lookin at its composition.

The anions also must be taken into account. Oxygen and fluorine have similar contributions. For chalcogenides, sulphur has a higher contribution to phonon energy tha Se and Te. In complex multicomponent glasses you should focus on lighter anions, and on cations that exhibit the larger "field strength".

Finally, as far as phonon energy is used to estimate the non radiative desexcitations in rare earth cations, you should also consider the local environnment of the rare earth. Unfortunately, it remains largely unknown. But it is possible to have some indirect information about it.

Last point: some impurities (hydroxyls) have also a significant influence.

Glass phonon energy is not a precise term but a rough estimation of the highest and significant vibrational transitions as explained above.

The maximum phonon energy can be attributed to the highest vibration energy of the glass forming cation bonding. It is a bulk property and defines the IR window of the glass. In case of complex mixed glasses, the highest significant vibrational peak.

There is another term called effective phonon energy which is usually used for estimation of deexcitation probability in rare earth doped glasses. This is related with the local environment as explained in the above comment. In phosphate glass composition, the maximum phonon energy is about 1250 cm-1 related to the vibrations of P=O bonding. However, the rare earth are not directly connected with this bond but with the second higher vibrations (P-O-) at about 1100 cm-1. Therefore the effective phonon energy is considered to be 1100 cm-1. 

Dr Marcel Poulain's suggestions  appear to be correct, of course under the assumption that " phonon spectrum" charactreizing a highly disordered system such as a complex glass is a valid proposition.

Thanks so much for your comments. I used to work with crystals and the "Glass world" is bit strange for me.

Finally for the identification of the highest phonon frequency, I used this ref: 10.1016/0022-3093(94)00648-2 (Raman studies in BGG glasses). They use the highest Raman active mode of the glass and it is compared with the alpha-GeO2 counterpart. I did something similar and I compared with DFPT simulations of their of each "constituent crystals". However, I am still have some doubts with that, if one of you have any reference or book to follow will be very helpful.

Best Regards E.

Atul D Sontakke Can you send any reference paper where this can be verified 

"In phosphate glass composition, the maximum phonon energy is about 1250 cm-1 related to the vibrations of P=O bonding. However, the rare earth is not directly connected with this bond but with the second higher vibrations (P-O-) at about 1100 cm-1. Therefore, the effective phonon energy is considered to be 1100 cm-1. "

Thanks for your comments. Finally for this work I found that the water inclusions in the glass. However, as you said the rare earth are not connected and this will be dependent on multiphonon transition probabilities. In the end, I didn't have time to calculate everything but it was a very nice experience with "phonon" in glasses.