The reason you want to add an electrolyte, is for conductivity. A good rule of thumb is the concentration should at least be 0.5 M to minimize Ohmic losses. Obviously, conductivity depends on the type of electrolyte.

A second consideration is pH. Some materials are only stable and/or active under certain conditions.

TiO2 is often used under acidic conditions, so you could use e.g. H2SO4 or HClO4. An alternative is to start out with dissolved salt (e.g. 0.5 M Na2SO4) and acidify with the acid.

If you want to work at neutral pH, your best option is a buffered solution (e.g. phosphate or borate), to minimize local pH fluctuations. For a quick test, you could also use something like KCl.

PS: are you sure you will use the TiO2 as a photocathode? It is probably n-type, and functions as a photoanode

Jan Rongé thank you! yes i want to use as photocathode; the electrode is hetero-junction.

You can use standard electrolytes, or if you want to increase your photocurrent and for instance your IPCE, go for europium compounds. Instead or reducing water, you will reduce Eu3+ to Eu2+, which is just one electron transfer and thus, easier to reduce 

Dear Andoshe

I am not sure if you are going to use TiO2 ans a cathode. Perhaps you are going to use TiO2 with sensitizer (to make it function in visible light).

TiO2 has wide band gap (~3.2 eV) which need wavelengths shorted than 400 nm (UV).

However, the redox couple is chosen based on the following:

1) Its potential should match well with the band edges of the semiconductor electrode.

2) Try to avoid redox couples that are corrosive or hazardous to environment

3) Think of stable systems with low cost

There are many different redox couple systems, aqueous and non-aqueous. If the aqueous system works well I recommend it for environmental reasons.

MOreover, different semiconductors need different redox couples. One needs to examine different redox couples to optimize one's system.  Just be careful