Heterogeneity in Local Chemical Bonding Explains Spectral Broadening in Quantum Dots with Cu Impurities

Abstract

Quantum dots (QDs) with optically active Cu impurities have been proposed as heavy-metal free alternatives to Cd and Pb chalcogenides. However, the origin of their unusual optical properties is not well understood. In particular, spectral broadening is an issue for their use in high color purity light-emitting diodes, and reabsorption-free solar windows. Here, we show with density functional theory calculations that chemical bonding variations have a major effect on the optical properties of Cu doped ZnSe QDs. The Cu-Se coordination sphere is highly covalent, and therefore sensitive to local variations in electrostatics and bond geometry. Correspondingly, changes in the Cu impurity environment lead to large shifts in their ground state energy, which causes spectral broadening when multiple Cu impurity bonding environments coexist as subensembles with distinct absorption and emission energies. We conclude that while electronphonon coupling is stronger for these systems than for typical II-VI QDs, spectral broadening predominantly occurs due to the inhomogeneous spatial distribution of Cu impurities. This is in agreement with studies that have shown narrow (~60 meV) single-particle emission linewidths in related Cu x In 2-x Se y S 2-y , or "CIS" QDs, which also emit through Cu impurities. Hence, we predict that narrow ensemble emission in photonic devices can be achieved if heterogeneity is controlled.

Spectroscopic and Magneto-Optical Signatures of Cu1+ and Cu2+ Defects in Copper Indium Sulfide Quantum Dots

“…This behavior can be ascribed to the effect of "spectroscopic size selection", which occurs due to excitation of progressively lager particles with a smaller bandgap from a polydisperse QD ensemble. 51 In the case of CIS QDs, however, this effect is weak as the PL ensemble line width is dominated not by QD size dispersion but by the distribution of the Cu-defect energies; 42,43 hence, the observed spectral shifts are small.…”

“…The peculiarities of CIS­(Se) QD photophysics include an unusually large Stokes shift between the photoluminescence (PL) band and the absorption onset (300–500 meV vs <100 meV in CdSe QDs), a strong PL broadening (>300 meV vs <100 meV in CdSe QDs), very long radiative lifetimes (100–500 ns vs ∼20 ns in CdSe QDs), and a featureless indirect-semiconductor-like absorption edge. These properties have been often ascribed to the involvement of optically active intragap defects. ,,,− In particular, key spectroscopic observations can be explained by a phenomenological model wherein the PL originates not from the intrinsic QD exciton recombination but instead involves a deep hole-like defect state, which optically couples to the band-edge electron. ,,− ,,,,− The surprising persistence of this defect across all reported CIS­(Se) QD samples indicates that it is native to these materials. Furthermore, the strong similarity between optical and magneto-optical properties of CIS­(Se) QDs and II–VI QDs intentionally doped with Cu ions suggests that this native defect is Cu-related, as was proposed in ref on the basis of a side-by-side comparison of CIS, Cu–doped ZnSe, and CdSe QDs.…”

“…Understanding the mechanisms underlying light–matter interactions in CIS­(Se) QDs is critical to ongoing efforts to rationally control their optical and electro-optical functionalities for existing and emerging applications. One especially important question is whether the unusual photophysical behaviors of CIS­(Se) QDs are intrinsic to these materials (as in the self-trapped exciton model , ) or if they are derived from crystal imperfections (as in the “intragap-defect” models ,,,,,,− ). Further, if these properties are intrinsic, then why are they so different from those of either parental I–III–VI bulk crystals or structurally similar II–VI CdS­(Se) QDs?…”

Stoichiometry-controllable optical defects in CuxIn2−xSy quantum dots for energy harvesting

“…11,51 However, the occurrence of this shift only at antisite defects, and the simultaneous presence of delocalized bandedges is still consistent with two-band transient absorption from "Cu 1+ " defects and narrow single particle emission (~60 meV) broadened by heterogeneity. 7,18,20,56 Our computational predictions imply that the performance of CIS QDs in energy harvesting applications can be improved by tuning the relative concentration of anti-site defects and copper vacancies through stoichiometry. LSCs and solid-state solar cells utilizing Cudeficient QDs should have reduced reabsorption and nonradiative recombination losses, and improved scaling of device performance.…”

“…S7), and ∆ S is impacted by the local defect bonding environment as expected by single-particle spectroscopy studies on CIS QDs, and DFT studies on related Cudoped ZnSe QDs. 18,56 Moreover, the band-edges are delocalized, and we expect both delocalized and localized transitions to occur.…”