Use fused quartz slides. Fused quartz will stand the heat treatment better. Some vendors will provide purity data.

It is necessary to investigate film substrate interactions as you have done. There are many surprises that occur at an interface. For example, a gold film on a silicon substrate can result in silicon diffusing to the surface of the gold to form  a thin layer of SiO2 at room temperature.

One way is to deposit a suitable diffusion barrier layer between the substrate and the thin film.

Thanks Syed Zameer. I heard stable insulator metal oxide such as SiO2 shall be okay. I used to prepare SiO2 layer using sol-gel solution derived from TEOS but not able to make it dense by thermal treatment for further than 650degC due to substrate limitation. I'm not able to test the porosity but refractive index showed it was comparable to silica film(n around 1.44xx). Any suggestion for type of material for diffusion barrier layer?

Use fused quartz slides. Fused quartz will stand the heat treatment better. Some vendors will provide purity data.

It is necessary to investigate film substrate interactions as you have done. There are many surprises that occur at an interface. For example, a gold film on a silicon substrate can result in silicon diffusing to the surface of the gold to form  a thin layer of SiO2 at room temperature.

Thank you for your suggestion Dr Alvarez.

Borosilicate glass is unstable for T>500°C - to my personal experience, it also starts to melt at slightly larger temperatures. We have worked with similar sample systems, and it turned out that it is preferable to use lower temperatures (500°C maximum) but increase the annealing time instead (to several hours). It is also favourable to keep the temperature at lower level in order to avoid formation of rutile, while the anatase appears to be more stable at lower temps. See attached paper, Good luck, Dirk

A nitride layer with good thermal stability may be helpful if deposited over Silica layer or directlty onto the glass slide before making a titania layer.

Hi, you could deposited SiO2 layer (barrier layer)  before final layer for two times.

Thanks for your suggestion. Apart from surface warping & elementals migration, I also treat TiO2 thin film around 500degC to avoid present of rutile. It good to know that other researchers using the same manner for the same reason. Thanks Dr Luetzenkirchen-Hecht for the nice paper on TiO2 crystallization. It will deepen my understanding.

Based on Prof Behzad suggestion, I think film thickness is a factor & should be consider for preventing elemental migration

Dear Mr. Tamizi, your suggestion about film thickness is Ok, but over optimum thickness results to cracking thin film during calcination. You must be attention to this subject at during work.

I understand there is an issue on thermal treatment for oxide layer

Thank you Prof. Just to continue.. I understand there is an issue on thermal treatment for oxide layer especially silica film derived from sol-gel coating technique. Used to fabricate silica-on-silicon(SOS) using solgel-spin coating-rapid thermal annealing technique for cladding layer on optical waveguides fabrication. The thickness is around 100nm per spin&anneal otherwise it will crack. It a pain-staking work to achieved non-crack of 2micron layer but it a nearly pure silica film. That was before I switch to organic-inorganic materials usually vinyl-silsesquioxane has been used as cladding layer. Normally, up to 2micron can be achieved by single spin- or dip-coating technique just the tricky part is how to make it dense.

I've dip-coated TiO2 film on vinyl-silsesquioxane film (as barrier layer). Thermal treatment has been carried out up to 300degC(will raise it to higher temp or prolong duration time). Just to have some comments on the photo-catalytic activities of the TiO2 film before undergo UV-Vis irradiation in methylene blue test. During photo-catalysis process, we understand there is a degradation process on the film surface that in contact with methylene blue solution but will there be a photo-catalysis process on the other side(I mean the surface between film that in contact with the barrier layer)?