Comment on "Self-Purification in Semiconductor Nanocrystals"

In a recent Letter [PRL 96, 226802 (2006)], Dalpian and Chelikowsky claimed that formation energies of Mn impurities in CdSe nanocrystals increase as the size of the nanocrystal decreases, and argued that this size dependence leads to "self-purification" of small nanocrystals. They presented density-functional-theory (DFT) calculations showing a strong size dependence for Mn impurity formation energies, and proposed a general explanation. In this Comment we show that several different DFT codes, pseudopotentials, and exchange-correlation functionals give a markedly different result: We find no such size dependence. More generally, we argue that formation energies are not relevant to substitutional doping in most colloidally grown nanocrystals.