The effect you mention is the Stokes shift. The shift comes from other secondary non-radiative deactivation pathways from the excited state. In quantum dots, due to the presence of surface states, there are more non-radiative deactivation pathways which result on a larger stokes shift.
Unfortunately, I will have to disagree with my colleague above.
The non-radiative de-activation pathways through surface traps create blinking, but are certainly not responsible for the Stokes shift observed in colloidal QDs (the word "colloidal" here is important, because, for example, InGaAs semiconduictor QDs, there is no Stokes shift).
The observed Stokes shift in colloidal QDs is due to the fact that the ground state is an F=2 state and thus is a dark state. Excitons on the first excited level thus cannot relax radiatively to F=2.
Thus, one can engineer the Stokes shift in colloidal QDs. For small ones, it is usually large (up to several tens of nm), while for large one, it gets smaller and tends to vanish.
The physical origin of Stokes shift is thus different than for Dye molecules, where it is due to the deformation/rotation energy of the molecule itself. It does not apply for colloidal QDs.
Okay, lets rephrase the answer. Quantum dots have other deactivation pathways as in Dye molecules. Quantum dots have energetic states within its band gap, which are intrinsic defects states, or can be introduced by doping. This intraband energetic levels can be radiative or non-radiative. The stokes shift comes from the emission of such radiative intraband energy levels. Depending on their position within the band gap, you get different stoke shifts.
Again, QDs, as in your reference, CdSe for instance, do not need to have defects in their bandgap to show a Stokes shift.
Even for a perfect QD, without any defect in its bandgap, their is a Stokes shift (several nm to several tens of nm), which is due to the ground state being an F=2 momentum state, which is dark. No need for defects, it is the ground state of the QD itself.
No doubt you can tune the Stoke shift with defects, but it is not the fundamental reason of the existence of Stokes shifts in colloidal QDs.