Quenching of the luminescence is often caused by effective energy loss through the surface effects, which is a consequence of the widespread of the wavefunction of your exciton. Surface layer of the material with higher energy gap should localize the wavefunction and thus increase the quantum yield. Of course lattice constant should fit.
Surface is indeed an defect state that can cause loss of the excitation. I would suggest to cover it with some other material in order to form a core-shell structure.
Earlier it was shown that for porous silicon radiative surface-related states in the form of ‘‘self-trapped excitons’’ at Si-Si dimers on the surface of nanocrystallites exist. They mostly exist on surfaces where the elastic response of the material is the weakest, and for crystallites as small as possible. In paper (PHYSICAL REVIEW B VOLUME 56, NUMBER 4 15 JULY 1997-II ) the various photoexcitation pathways involved in accessing and populating these molecular states were examined. Absorption and fluorescence spectra and the quantum efficiency of the photoluminescence were determined. Obtained results demonstrates a dramatic enhancement in the efficiency for crystallite sizes below a critical size near 1.4 nm.
I can not say that these results are relevant to cadmium telluride. However, the role of surface course plays an important role in photoluminescence efficiency of QDs.