any transition inside one electronic subshell is forbidden by Laporte rule. In case of free metal ion in vacuum d-d transitions are very weak. When the ion is embedded into the matrix with the low order of crystal field symmetry at the ion site, d-electrons change their trajectories and cannot be described by pure d-state wavefunctions. Their wavefunctions often involve odd parity state members so the mixed ground and excited states are formed and transitions between them are partially forbidden, hence the intensity for these transitions are much higher then for free ion in vacuum. The valent interaction also has impact on the degree of the mixing. For ionic interaction the mixing is stronger and the luminescence should be more intensive then for covalent one.
Dear Anatoly P. Pusharev, Thank You for reply. Your answer very helping.can you please elaborate the sentence " For ionic interaction the mixing is stronger and the luminescence should be more intensive then for covalent one ".
I have calculated Racah parameters for highly luminescent nanocrystals of Mn doped ZnSe. From the B parameter (inter d-electron repulsion parameter ) i have observed the increased covalent interaction between metal and ligand.
In my opinion the increased covalent interaction leads to decrease in intensity of d-d transitions as the d-p mixing is reduced. Furthermore, I think you can check your findings looking at charge transfer (CT) band in absorption. Increase in covalency should result in redshift of this band with respect to that of material in which the less covalent interaction was found.