Autonomous robotic exploration of deep-blue perovskite NCs achieving near-unity PLQY
Jongyeon Kim a, Yanqiu Jiang a, Bartosz Grzybowski a b
a Center for Algorithmic and Robotized Synthesis (CARS), Institute for Basic Science (IBS)
b Department of Chemistry, Ulsan National Institute of Science and Technology (UNIST)
Proceedings of Hybrid and Perovskite materials for energy, lighting, sensing and computing (HYPE26)
Athens, Greece, 2026 June 22nd - 24th
Organizers: Maria Vasilopoulou and Thomas Stergiopoulos
Poster, Jongyeon Kim, 029
Publication date: 15th May 2026

Navigating the vast, high-dimensional chemical space of colloidal nanocrystals remains a formidable challenge, as non-additive compositional and surface interactions render one-variable-at-a-time experimentation fundamentally insufficient for rational design with targeted optical properties. Autonomous, closed-loop synthesis platforms have emerged as powerful tools to address this, yet existing approaches are limited by optimizing crude reaction mixtures whose optical properties bear little resemblance to purified, processed material.

Here we present a closed-loop robotic platform based on our house-built Robowski [1,2], uniquely integrating automated synthesis, purification, post-synthetic passivation, and photophysical characterization within a single workflow, producing fully processed nanocrystals at a throughput exceeding 200 formulations per day. We apply this platform to the targeted discovery of highly photoluminescent deep blue-emitting mixed CsPbBr3-xClx nanocrystals (λem < 465 nm), the most demanding color band for perovskite light-emitting diodes.

Robowski synthesizes monodisperse spheroidal CsPbBr3 NCs via ligand-assisted nucleation at room temperature [3], followed by metal chloride exchange to blue-shift emission toward the target range. Twelve metal chlorides are systematically screened, seeding a closed-loop Bayesian optimization campaign, in which each cycle the platform synthesizes suggested compositions, acquires their photophysical data, and feeds the results back to the model to guide the next batch.

Navigating through a theoretical space of ~100,000 multi-metal co-dopant compositions, just 278 experiments across four iterative cycles successfully uncover a Zn-dominant domain as the most photoluminescent region. Subsequent exploration across ten post-synthetic passivation agents spanning inorganic cations, small organic ammoniums, and long-chain ammonium salts reveals that a synergistic combination of primary and quaternary long-chain ammonium salts achieves surface reconstruction inaccessible to either agent alone.


Manual validation of the champion formulation yields a near-unity photoluminescence quantum yield of 94.7% at λem = 458 nm even after three washing cycles, compared with 14.9% for untreated sample. Ranking among the highest reported PLQYs for deep blue-emitting perovskite nanocrystals, this work demonstrates the potential of accelerated, closed-loop robotic platforms as a powerful and enabling strategy for the targeted discovery of high-performance functional nanomaterials.

© FUNDACIO DE LA COMUNITAT VALENCIANA SCITO
We use our own and third party cookies for analysing and measuring usage of our website to improve our services. If you continue browsing, we consider accepting its use. You can check our Cookies Policy in which you will also find how to configure your web browser for the use of cookies. More info