2024 AIChE Annual Meeting

(690a) Material-Efficient Self-Driving Lab for Multi-Objective Parameter Space Mapping of Inorganic Materials

Authors

Nikolai Mukhin - Presenter, North Carolina State University
Fernando Delgado Licona, North Carolina State University
Sina Sadeghi, Texas Tech University
Milad Abolhasani, NC State University
Metal halide perovskite (MHP) quantum dots (QDs) have demonstrated promise to revolutionize fields such as displays, light-emitting diodes, photovoltaics, and quantum information science. Known for their high absorption coefficient, long diffusion length, and excellent charge transport, MHP QDs offer superior performance compared to other classes of semiconductor materials. With high photoluminescence quantum yield (PLQY) and narrow emission linewidth, MHP QDs provide tunable optical and optoelectronic properties through size, morphology, and composition adjustments [1]. Despite the proven potential of MHP QDs, their synthesis and development are pursued through time- and labor-intensive manual batch experimentation.

Machine learning (ML)-guided automated experimentation in a closed-loop format known as self-driving labs (SDLs) has recently gained popularity in chemical and materials sciences for its use of experimental planning, parameter space navigation with ML modeling, and decision-making techniques that can be used for complex/high dimensional reactions [2]. When multiple output parameters compete with one another (for example minimizing emission linewidth and maximizing PLQY of MHP QDs), identifying the Pareto-front and pushing its boundary is crucial for finding the optimal synthetic routes of emerging functional materials.

Flow chemistry has been demonstrated to adapt well to the fast formation kinetics of MHP QDs and even has additional benefits through continuous experimentation, in-situ characterization, and reaction miniaturization (lower chemical consumption and waste generation than batch experimentation), which makes it a great reactor of choice for SDLs [3]. However, the current waiting times of continuous flow reactors to reach steady-state operation hampers the utilization and widespread adoption of this powerful research acceleration tool in chemical and materials sciences.

In this work, we present a material-efficient SDL equipped with a single-droplet fluidic platform, specifically designed for rapid synthesis science studies of MHP QDs. Following the characterization, validation, and benchmarking of both the hardware and ML agent of the developed SDL, we employ this powerful autonomous experimentation technology for accelerated multi-objective synthesis space mapping of cesium lead halide perovskite QDs with 5 uL droplets per reaction condition. Next, we demonstrate the power of the developed material-efficient SDL for autonomous synthetic route discovery of high-performing MHP QDs with the highest PLQY with the narrowest emission linewidth at a wide range of the UV-Vis spectrum through closed-loop Pareto-front mapping.

References:

[1] Protesescu, L.; Yakunin, S.; Bodnarchuk, M. I.; Krieg, F.; Caputo, R.; Hendon, C. H.; Yang, R. X.; Walsh, A.; Kovalenko, M. V. Nanocrystals of Cesium Lead Halide Perovskites (CsPbX 3 , X = Cl, Br, and I): Novel Optoelectronic Materials Showing Bright Emission with Wide Color Gamut. Nano Lett. 2015, 15 (6), 3692–3696. https://doi.org/10.1021/nl5048779.

[2] Abolhasani, M.; Kumacheva, E. The Rise of Self-Driving Labs in Chemical and Materials Sciences. Nat. Synth. 2023, 2 (6), 483–492. https://doi.org/10.1038/s44160-022-00231-0.

[3] Bateni, F.; Sadeghi, S.; Orouji, N.; Bennett, J. A.; Punati, V. S.; Stark, C.; Wang, J.; Rosko, M. C.; Chen, O.; Castellano, F. N.; Reyes, K. G.; Abolhasani, M. Smart Dope: A Self-Driving Fluidic Lab for Accelerated Development of Doped Perovskite Quantum Dots. Adv. Energy Mater. 2024, 14 (1), 2302303. https://doi.org/10.1002/aenm.202302303.