CdTe quantum dots have unique characteristics that are promising for applications in photoluminescence, photovoltaics or optoelectronics. However, wide variations of the reported quantum yields exist and the influence of ligand-surface interactions that are expected to control the excited state relaxation processes remains unknown. It is important to thoroughly understand the fundamental principles underlying these relaxation processes to tailor the QDs properties to their application. Here, we systematically investigate the roles of the surface atoms, ligand functional groups and solvent on the radiative and non-radiative relaxation rates. Combining a systematic synthetic approach with X-ray photoelectron, quantitative FT-IR and time-resolved visible spectroscopies, we find that CdTe QDs can be engineered with average radiative lifetimes ranging from nanoseconds up to microseconds. The non-radiative lifetimes are anticorrelated to the radiative lifetimes, although they show much less variation. The density, nature and orientation of the ligand functional groups and the dielectric constant of the solvent play major roles in determining charge carrier trapping and excitonic relaxation pathways. These results are used to propose a coupled dependence between hole-trapping on Te atoms and strong ligand coupling, primarily via Cd atoms, that can be used to engineer both the radiative and non-radiative lifetimes.

Currently, the most common way to reduce blinking in quantum dots (QDs) is accomplished by using very thick and/or perfectly crystalline CdS shells on CdSe cores. Ideally, a nontoxic material such as ZnS is preferred to be the outer material in order to reduce environmental and cytotoxic effects. Blinking suppression with multishell configurations of CdS and ZnS has been reported only for "giant" QDs of 15 nm or more. One of the main reasons for the limited progress is that the role that interfacial trap states play in blinking in these systems is not very well understood. Here, we show a "Goldilocks" effect to reduce blinking in small (∼7 nm) QDs by carefully controlling the thicknesses of the shells in multishell QDs. Furthermore, by correlating the fluorescence lifetime components with the fraction of time that a QD spends in the on-state, both with and without applying a threshold, we found evidence for two types of blinking that separately affect the average fluorescence lifetime of a single QD. A thorough characterization of the time-resolved fluorescence at the ensemble and single-particle level allowed us to propose a detailed physical model involving both short-lived interfacial trap states and long-lived surface trap states that are coupled. This model highlights a strategy of reducing QD blinking in small QDs by balancing the magnitude of the induced lattice strain, which results in the formation of interfacial trap states between the inner shell and the outer shell, and the confinement potential that determines how accessible the interfacial trap states are. The combination of reducing blinking while maintaining a small overall QD size and using a Cd-free outer shell of ZnS will be useful in a wide array of applications, particularly for advanced bioimaging.

No abstract

Coordinating ligands are widely used to vary the solubility and reactivity of nanoparticles for subsequent bioconjugation. Although long-term colloidal stability is enhanced by using bidentate coordinating ligands over monodentate ones, other properties such as non-specific adsorption of target molecules and ligand exchange have not been quantified. In this study, we modified a near infrared dye to serve as a highly-sensitive reporter for non-specific binding of thiolated target molecules to nanoparticle surfaces that are functionalized with monodentate or bidentate coordinated ligands. Specifically, we analyzed non-specific binding mechanisms to quantum dots (QDs) by fitting the adsorption profiles to the Hill equation and the parameters are used to provide a microscopic picture of how ligand density and lability control non-specific adsorption. Surprisingly, bidentate ligands are worse at inhibiting adsorption to QD surfaces at low target:QD ratios, although they become better as the ratio increases, but only if the nanoparticle surface area is large enough to overcome steric effects. This result highlights that a balance between ligand density and lability depends on the dentate nature of the ligands and controls how molecules in solution can coordinate to the nanoparticle surface. These results will have major implications for a range of applications in nanobiomedicine, bioconjugation, single molecule spectroscopy, self-assembly and nano(photo)catalysis where both non-specific and specific surface interactions play important roles. As an example, we tested the ability of monodentate and bidentate functionalized nanoparticles to resist non-specific adsorption of IgG antibodies that contained free thiol groups at a 1:1 QD:IgG ratio and found that QDs with monodentate ligands did indeed result in lower non-specific adsorption.

Choosing the composition of the shell for QDs is not trivial, since both the band-edge energy offset and interfacial lattice mismatch play roles in influencing the final optical properties. One way to balance these competing effects is by forming multi-shells and/or gradient-alloyed shells. However, this introduces multiple interfaces and their relative effects on quantum yield and blinking are not yet fully understood. Here we undertake a systematic, comparative study of adding inner shells of single composition vs gradient-alloyed shells of cadmium/zinc chalogenides onto CdSe cores, and then capping with a thin ZnS outer shell to form various core/multi-shell configurations. We show that the inner shell architecture between the CdSe core and the outer ZnS shell plays a significant role in both quantum yield and blinking dynamics but that these effects are not correlated – a high ensemble quantum yield doesn’t necessarily equate to reduced blinking. Two mathematical models have been proposed to describe the blinking dynamics – the more common power-law model and a more recent multi-exponential model. By binning the same data with 1 ms and 20 ms resolution, we show that the on-times can be better described by the multi-exponential model while the off-times can be better described by the power-law model. We discuss physical mechanisms that might explain this behavior and how it can be affected by the inner shell architecture.