Synthesis and post-synthetic modifications of CuInS2 nanocrystals

Abstract

Copper chalcogenide nanocrystals (NCs) are of great interest as replacements for the highly researched, but toxic cadmium and lead based chalcogenide nanocrystals due to their low toxicity, environmental compatibility, potentially lower costs and very wide range of compositions and crystal structures. CuInS2 is a ternary copper chalcogenide, where the photoluminescence of CuInS2 nanocrystalsspans the red to near-infrared (NIR) region. The photoluminescence quantum yield (PLQY) of CuInS2 nanocrystals is typically below 5-10% and must be improved for efficient use in optical applications like white LEDS, luminescent solar concentrators and bioimaging. Improving the PLQY is generally achieved by overcoating the nanocrystal with a wider bandgap semiconductor material, preferably without cadmium. ZnS is a good alternative compared to cadmium based shells due to its small lattice mismatch with CuInS2 ( ̴2.35%) and lower toxicity. The problem of using ZnS as shell material is the possibility of interdiffusion of Zn2+ into the CuInS2 core, leading to blue-shifts and broadening in optical spectra. For applications like bioimaging this can be an issue where a specific spectral range is used. To obtain high PLQY CuInS2/ZnS nanocrystals without alloying, a gallium-rich (Cu,In,Ga)S2 layer in between could prevent interdiffusion of Zn2+ into the CuInS2 core as the ionic radius of Ga3+ is different from Cu+, In3+ and Zn2+. This would hamper the blue-shift and broadening of the optical spectra. The (Cu,In,Ga)S2 layer could be obtained by a Ga3+ for Cu+ cation exchange with GaCl3-diphenylphosphine (DPP) as the precursor. The changes in PLQY, absorption and photoluminescence emission (PL) spectra of bare CuInS2 during the heating-up synthesis with varying volumes of 1-dodecanethiol (DDT) are followed. Optical spectra red-shift with increasing reaction times due to nanocrystal growth and the PLQY increases with increasing reaction times likely due to reduction in defect density, the decrease of the surface/volume ratio and/or composition changes. The influence of variations in synthesis parameters, such as concentration and volume of GaCl3-DPP, temperature, time, and solvent on the Ga3+ for Cu+ cation exchange in CuInS2 nanocrystals are investigated to gain insight into the reaction. Blue-shifts in the order of 10-102 meV are observed in PL spectra of the nanocrystals after the reaction with diluted and neat GaCl3-DPP. Absorption spectra become better defined, likely due to etching, and PL increases are observed only when neat GaCl3-DPP is used. EDS analysis and elemental mapping confirms presence of gallium, but do not show agreement with elemental composition changes that should preserve charge balance in Ga3+ for Cu+ exchange. As PL spectra are influenced in both the dilute and concentrated regime of precursor whereas the absorption spectra are not, it is proposed that a gallium complex acts as a ligand which only affects the excited state due to electron repulsion between the exciton and the gallium complex, resulting in blue-shifted PL spectra. This could also explain PL intensity enhancements due to surface passivation by the complex and the larger than expected concentration of gallium found on the nanocrystals with EDS. However, it is still speculative and uncertain how gallium is bound to the nanocrystal surface. As the formation of a (Cu,In,Ga)S2 shell on the CuInS2 nanocrystals appeared to be unsuccessful, a shelling reaction with ZnS was performed on bare CuInS2 nanocrystals. This resulted in nanocrystals with a higher PLQY, but blue-shifted and broadened optical spectra, attributed to alloying. Further research is needed to examine how gallium is bound to the nanocrystals in this work. It should also be investigated whether a (Cu,In,Ga)S2 shell is formed and if this indeed can prevent interdiffusion.