(17b) Multi-Stage Microfluidic Growth and Shelling of Quantum Dots

Authors: 
Abolhasani, M., Massachusetts Institute of Technology
Hassan, Y., University of Toronto
Kumacheva, E., FlowJEM Inc.
Scholes, G. D., University of Toronto
Guenther, A., University of Toronto



Quantum dots (QDs) demonstrate a wide range of desirable size- and shape-dependent properties that benefit their application as probes for biological imaging in vitro and in vivo, in molecular diagnostics and for device-level integration (e.g., in light emitting diodes, displays, solar cells and sensors). While available for more than 20 years, the solution-phase processes commonly employed for the batch-scale preparation of QDs lack precise control of reaction conditions as well as scalability, and the spectral characteristics of as-prepared nanocrystals which frequently display unwanted batch-to-batch variations. Microfluidic (MF)-based technologies have already demonstrated promising solutions to overcome some of the aforementioned challenges. However, significant hurdles remain. For instance, at a given throughput, the required microreactor footprint and cost increase exponentially with the increasing target peak emission wavelength.

We present an automated MF approach for the routine preparation of high quality CdSe and CdS QD cores, with a full width at half maximum (FWHM) emission peak wavelength smaller than 32 nm and a wide range of peak emission wavelengths (450-650 nm). Our approach takes advantage of a unique multi-pass microreaction platform. The employed MF device was fabricated using bulk silicon micromachining procedures. Two heating zones were accommodated on the silicon-microfabricated microreaction platform to deterministically initiate and stop nanocrystal growth during each pass. We demonstrate the predictive stepwise growth of large QDs (i.e., ~5-6 nm size, PL peak ~ 600-650 nm) to accommodate for the long required residence times of 200s - 400s.  The stepwise growth approach allows adding inorganic shells of either CdS or ZnS to as-prepared QD cores. As a result, the number of surface defects is reduced and the quantum efficiency predictively increases as compared to previously synthesized QD cores while the FWHM remained unaltered.