Which temperature did you fabricate the film? If you take the sample before self-cooling , it can crack! If ou wait for -self-cooling, may be you can get a good film. This is my suggestion.
I agree with Ali, thin films are highly sensitive to external stress or strain. A small stress can lead to nano or micro crack formation over your film. If you avoid thermal stress generation in silicon, you can get crack-free thin films.
TiO2 films deposited onto the silicon substrate at 2000 rpm, 50 sec, gives good surface coverage provided the substrate is plasma cleaned for 2 min. Try to dry clean the substrate before you coat the film. Coming to annealing temperature try to ramp 10 deg per min towards 500 deg C. Once you soak at 500 C for 30 min allow the furnace to cool naturally, leave it over night, hope then you may avoid the cracks. The approximate thickness of the film would be 40 nm to 50 nm. Hope you got the answer..
Its a very common issue for thin films, because of different Thermal Expansion Coefficients (TCE) of film and substrate, which leads to strain and stress and ultimately nano and micro cracks developments in the film.
In order to avoid cracks in films, need to control substrate temperature, deposition rate, and cooling or annealing rate.
If you give the deposition method and conditions of your films, you will get better solution for your problem.
More information about the deposition method is definitely required. Nevertheless, it could be helpful to deposit a thin intermediate layer of Ti between the substrate and the oxide layer.