I think this is may be the effect of restructuring and ordering the dislocation structure. Dislocations can arise because the lattice constant of Co differs from the lattice constant TiO2 (and may be the lattice constant of used substrate). Dislocations are collected during annealing and form the sub-boundaries. In result the grain structure are crushed.
Annealing may has positive or negative effect to the material in the form of bulk or thin film.
This depends on the bond strength and bond energy. If bond strength and bond energy had lower values so the annealing process may has a negative effect and broke some of these bonds to be dangling or unsaturated bond, such this issue increases the concentration or the No. of wrong bonds (dangling or unsaturated bond) which increase the disorder of the material. Also annealing may has a positive effect on the material, the annealing temperature may enhance the order through the material according to it's value; annealing makes the dangling bonds (imperfections) annealed out such this issue increases the order through the material. To check the suitable annealing temperature you need to process DTA or DSC analysis.