Copper's Role in the Photoluminescence of Ag1-xCuxInS2 Nanocrystals, from Copper-Doped AgInS2 (x ∼ 0) to CuInS2 (x=1)

A series of Ag1-xCuxInS2 nanocrystals (NCs) spanning from 0 <= x <= similar to 1 was synthesized by partial cation exchange to identify copper's contributions to the electronic structure and spectroscopic properties of these NCs. Discrete midgap states appear above the valence band upon doping AgInS2 NCs with Cu+ (small x). Density functional theory calculations confirm that these midgap states are associated with the 3d valence orbitals of the Cu+ impurities. With increasing x, these impurity d levels gradually evolve to become the valence-band edge of CuInS2 NCs, but the highest-occupied orbital's description does not change significantly across the entire range of x. In contrast with this gradual evolution, Ag1-xCuxInS2 NC photoluminescence shifts rapidly with initial additions of Cu+ (small x) but then becomes independent of x beyond x > similar to 0.20, all the way to CuInS2 (x = 1.00). Data analysis suggests small but detectable hole delocalization in the luminescent excited state of CuInS2 NCs, estimated by Monte Carlo simulations to involve at most about four copper ions. These results provide unique insights into the luminescent excited states of these materials and they reinforce the description of CuInS2 NCs as "heavily copper-doped NCs" in which photogenerated holes are rapidly localized in copper 3d-based orbitals.