you need put alcohol like isopropanol, NH3 ac, and the precursor vanadium isopropoxide..

Dear Christian Javier Salas-Juárez, do you have some papers on the topic. Could you give me more details about the recepie. What is the efect of ammonia?  I do not use it and obtain stable sols. Thank you

Vanadium(V) trisisopropoxide oxide (VTIP), can hydrolyse as:

2VO[(CH3)2CHO]3 + 3H2O → 6(CH3)2CHOH + V2O5

or, alternatively as:

VO[(CH3)2CHO]3 + 3H2O → 3(CH3)2CHOH + VO(OH)3

Eventually, being a relatively strong oxidant, vanadyl(V) trihydroxy can (partially) oxidise the formed isopropyl alcohol to acetone:

2VO(OH)3 + (CH3)2CHOH → 2VO2 + 4H2O + (CH3)2CO

It thus seems clear that if V2O5 is to be favoured, H2O should not be added in excess of a 3/2 H2O:VTIP molar ratio, while added acetone may contribute to avoid unwanted progress of last reaction.

It is as films

Thank you very much Dr. Carlos Araújo Queiroz! I will try to use acetone. Just to mention that it is not possible to use water with VTIP because even very small amounts lead to precipitation. Till now we have obtained  clear sol only using isopropanol alchohol. Even the usage of ethanol leads to the same effect as using water. The problem is that the deposited thin films have a relevent absorption that lead us to the conclusion that mixed oxides exist.

Dear Tsvetanka Babeva. I have prepared V2O5 films by sol-gel method using vanadium isopropoxide dissolved in the Isopropanol. the solution is stabilised with acetyl acetone (ACAC). The solution was stirred for 2 hours to be stabilized.

tHE THIN layers WAS DEPOSITED ON TO the silicon substrats  and  heated at 500°C durin 2 h to obtain nanarods of V2O5.

Dear Ayouz-Chebout Katia do you have some measurements proving that you obtain only V2O5 instead of some mixed sub oxides? Do you have some optical measurements? I am particularly interested in extinction coefficient in visible spectral range. Is the annealing performed in air?